Soft Open Points for the Operation of Medium Voltage ...s PhD thesis_02.02.2016 (clean... · Soft...

Soft Open Points for the Operation of

Medium Voltage Distribution

Networks

Wanyu Cao

School of Engineering

Cardiff University

A thesis submitted for the degree of

Doctor of Philosophy

October, 2015

ii

Acknowledgement

I would like to express my heartfelt gratitude to my supervisor, Prof. Jianzhong Wu, for

being the source of inspiration, advice and encouragement throughout my research

study. In addition, his patience and continuous guidance in improving my writing

ability is of great value to me and my future academic life.

I would like to express my sincere gratefulness to my supervisor Prof. Nick Jenkins,

CIREGS’ leader, for his invaluable assistance, guidance and constructive criticism in

conducting my research and developing this thesis.

I would like to thank Dr. Jun Liang and Sheng Wang for their kind support and

suggestions during the research study, especially in the power electronics filed.

I would like to thank Dr. Chao Long, Dr. Cheng Meng and Dr. Lee Thomas who

proofread my thesis and also suggested many improvements to the content.

I thank Cardiff University for giving the opportunity to pursue my research studies for a

PhD. I thank, North China Electric Power University, Cardiff University, EPSRC and

OPEN project, UK for funding my research study.

I thank all my friends and colleagues who helped in many ways during my stay in

Cardiff.

Finally, I would like to express my deepest thanks to my family for their endless

encouragement, understanding, patience and support all the while.

iii

Declaration

This work has not previously been accepted in substance for any degree and is not

concurrently submitted in candidature for any other higher degree.

Signed:……………………………..(Candidate) Date:………………………….

Statement 1

This thesis is being submitted in partial fulfilment of the requirements for the degree

of ……………..(insert as appropriate PhD, MPhil, EngD)


Statement 2

This thesis is the result of my own independent work/investigation, except where

otherwise stated. Other sources are acknowledged by explicit references.


Statement 3

I hereby give consent for my thesis, if accepted, to be available for photocopying,

inter-library loan and for the title and summary to be made available to outside

organisations.


iv

Copyright

Copyright in text of this thesis rests with the Author. Copies (by any process) either in

full, or of extracts, may be made only in accordance with instructions given by the

Author and lodged in the Library of Cardiff University. Details may be obtained from

the Librarian. This page must form part of any such copies made. Further copies (by

any process) of copies made in accordance with such instructions may not be made

without the permission (in writing) of the Author.

The ownership of any intellectual property rights which may be described in this thesis

is vested with the author, subject to any prior agreement to the contrary, and may not be

made available for use by third parties without his written permission, which will

prescribe the terms and conditions of any such agreement.

v

Abstract

Soft Open Points (SOPs) are power electronic devices installed in place of

normally-open points in electrical power distribution networks. They are able to

provide active power flow control, reactive power compensation and voltage regulation

under normal network operating conditions, as well as fast fault isolation and post-fault

supply restoration under abnormal conditions. The use of SOPs for the operation of

medium voltage (MV) distribution networks was investigated. Three aspects were

studied, which include the control of an SOP, benefit analysis of using SOPs and

distribution network voltage control with SOPs.

Two control modes for the operation of an SOP, which is based on back-to-back voltage

source converters (VSCs), were developed. The operating principle and performance of

the back-to-back VSC based SOP under both normal and abnormal network operating

conditions were analysed. It was found that during the change of network operating

conditions, smooth transitions between the two controls modes were needed. Using soft

cold load pickup and voltage synchronization processes can achieve the smooth mode

transitions.

A steady state analysis framework to quantify the operational benefits of a MV

distribution network with SOPs was developed, which considers feeder load balancing,

power loss minimization and voltage profile improvement. The framework also

considered traditional network reconfiguration and the combination of both SOP

control and network reconfiguration to quantify the benefits. It was found that in the

case study using only one SOP can achieve a similar improvement in network operation

compared to the case of using network reconfiguration with all branches equipped with

remotely controlled switches. The combination of both SOP control and network

reconfiguration can achieve the optimal network operation.

vi

A coordinated voltage control strategy for active distribution networks with SOPs was

developed considering the control of SOP, the on-load tap changer (OLTC) and

distributed generation (DG) units. Multiple objectives were considered, which include

maintaining network voltages within the specified limits, mitigating DG active power

curtailment, reducing the tap operations of the OLTC and the network power losses.

The proposed control strategy is based on a distributed control framework where each

control device is considered as a local control agent. A priority-based coordination

method is applied on the local control agents to obtain a trade-off among the objectives

based on their prioritisation. It was shown that, comparing with centralized control

strategies, the proposed control strategy can provide reliable distribution network

voltage control with less communication investments, reduced computation burden and

fewer data exchanges. It was also found that the SOP played a key role of

compensating for the OLTC control, avoiding unnecessary DG active power

curtailment, reducing the number of tap operations and network power losses as well as

reducing the amount of reactive power support from DG units for voltage control.

vii

Contents

Acknowledgement .......................................................................................................... ii

Declaration ..................................................................................................................... iii

Copyright ....................................................................................................................... iv

Abstract ........................................................................................................................... v

Contents ........................................................................................................................ vii

List of Figures ................................................................................................................ xi

List of Tables ................................................................................................................. xv

Nomenclature ............................................................................................................... xvi

Chapter 1 - Introduction................................................................................................ 1

1.1 Background ............................................................................................................ 2

1.1.1 Climate Change and Renewable Generation .................................................. 2

1.1.2 Electricity Demand Growth ............................................................................ 4

1.1.3 Emergence of Smart Grid Concept ................................................................. 5

1.2 Research Motivation .............................................................................................. 7

1.2.1 Control of an SOP ........................................................................................... 9

1.2.2 Benefits analysis of using SOPs ...................................................................... 9

1.2.3 Distribution network voltage control with SOPs .......................................... 10

1.3 Objectives and Contributions of this Thesis ........................................................ 11

1.4 Thesis Outline ...................................................................................................... 12

Chapter 2 - Literature Review .................................................................................... 14

2.1 Challenges for Distribution Network Operation .................................................. 15

2.1.1 Aging Assets and Lack of Circuit Capacity .................................................. 15

2.1.2 Operational Constraints ................................................................................. 16

viii

2.1.3 Reliability and Efficiency of Supply ............................................................. 17

2.2 Network Operation Functions .............................................................................. 18

2.2.1 Network Reconfiguration .............................................................................. 19

2.2.2 Coordinated Voltage Control ......................................................................... 20

2.2.3 Operation Functions Offered by Power Electronic devices .......................... 25

2.3 Soft Open Points .................................................................................................. 27

2.3.1 Benefits of Soft Open Points ......................................................................... 27

2.3.2 Types of Soft Open Points ............................................................................. 28

2.3.3 Previous Soft Open Points Studies ................................................................ 31

2.4 Summary .............................................................................................................. 33

Chapter 3 - Control of a Back-to-Back VSC based SOP .......................................... 35

3.1 Introduction .......................................................................................................... 36

3.2 Back-to-Back VSC based SOP ............................................................................ 36

3.3 Control Modes of the Back-to-Back VSC based SOP ......................................... 38

3.3.1 Power Flow Control Mode ............................................................................ 38

3.3.2 Supply Restoration Mode .............................................................................. 40

3.4 SOP Operation in MV Distribution Networks ..................................................... 42

3.4.1 Normal Conditions ........................................................................................ 43

3.4.2 During Fault Conditions ................................................................................ 45

3.4.3 Post-fault Supply Restoration Conditions ..................................................... 49

3.5 Summary .............................................................................................................. 57

Chapter 4 - Benefits Analysis of SOPs for MV Distribution Network Operation .. 59

4.1 Introduction .......................................................................................................... 60

4.2 Modelling of Soft Open Points ............................................................................ 60

ix

4.2.1 Physical Limitations of Back-to-Back Converters ........................................ 62

4.2.2 Internal Power Losses of Back-to-Back Converter ....................................... 63

4.3 Optimal Operation of the Soft Open Points ......................................................... 65

4.3.1 Problem Formulation .................................................................................... 65

4.3.2 Method of Determining Optimal SOP Operation ......................................... 67

4.4 Network Reconfiguration Considering SOPs ...................................................... 73

4.5 Case Study ............................................................................................................ 74

4.5.1 Improve Network Performances Using SOPs ............................................... 75

4.5.2 Improve Network Performance Considering both SOP and Network

Reconfiguration ...................................................................................................... 78

4.5.3 Impact of DG Connections............................................................................ 82

4.5.4 Impact of the SOP Device Losses ................................................................. 85

4.6 Summary .............................................................................................................. 87

Chapter 5 - Voltage Control in Active Distribution Networks with SOPs .............. 89

5.1 Introduction .......................................................................................................... 90

5.2 Voltage Control Framework ................................................................................. 90

5.2.1 Voltage Profile Estimation ....................................................................... 91

5.2.2 Proposed Voltage Control Framework .......................................................... 96

5.3 Coordinated Voltage Control Strategy ................................................................. 99

5.3.1 Effectiveness of SOP on Voltage Control ..................................................... 99

5.3.2 Priority-Based Coordination ....................................................................... 105

5.3.3 Implementation procedures of the information-sharing platform ............... 115

5.4 Discussion: Benefits of the Proposed Control Strategy ..................................... 116

5.5 Case Study .......................................................................................................... 117

x

5.5.1 Voltage Profile Estimation .......................................................................... 119

5.5.2 Coordinated voltage control using OLTC and SOP .................................... 121

5.5.3 Coordinated voltage control using OLTC, SOP and DG ............................ 125

5.6 Summary ............................................................................................................ 130

Chapter 6 - Conclusions and Future Work .............................................................. 132

6.1 Conclusions ........................................................................................................ 133

6.1.1 Control of a Back-to-Back VSC based SOP ............................................... 133

6.1.2 Benefit Analysis of SOPs for MV Distribution Network Operation ........... 134

6.1.3 Voltage Control in Active Distribution Networks with SOPs ..................... 135

6.2 Future ................................................................................................................. 136

6.2.1 Use of other control strategies for the operation of the back-to-back VSC

based SOP ............................................................................................................ 136

6.2.2 Performances of the back-to-back VSC based SOP considering different

load types ............................................................................................................. 137

6.2.3 Impacts of SOPs on feeder automation ....................................................... 137

6.2.4 Use of SOPs in unbalanced three-phase distribution networks ................... 138

6.2.5 Economic analysis of SOPs in MV distribution networks .......................... 138

Reference ..................................................................................................................... 140

Publications ................................................................................................................. 148

Appendix A: Data for the Two-Feeder Test Network ............................................. 149

Appendix B: Data for the Four-Feeder Test Network ............................................ 151

xi

List of Figures

Figure 1.1: The UK Renewable Energy Target [5] .......................................................... 3

Figure 1.2: World Electricity Consumption by Region [9] .............................................. 4

Figure 1.3: Critical developments for UK smart grid routemap out to 2050 [17] ........... 7

Figure 2.1: Three-feeder example network: (a) before network reconfiguration; (b) after

network reconfiguration [27] ......................................................................................... 19

Figure 2.2: Cascaded control architecture for centralized coordination [48] ................. 23

Figure 2.3: Coordination among three control agents via two ways communication [41]

........................................................................................................................................ 25

Figure 2.4: Simple distribution network: option A representing a NOP connection and B

representing an SOP [67] ............................................................................................... 27

Figure 2.5: Topologies of different types of SOPs [70] ................................................. 29

Figure 3.1: (a) Basic configuration of a distribution network with an SOP; (b) Main

circuit topology of the back-to-back VSC based SOP. .................................................. 37

Figure 3.2: Control block diagram of the SOP for power flow control mode: (a) outer

power control loop; (b) inner current control loop; (c) PLL controller ......................... 40

Figure 3.3: Control block diagram of the interface VSC for supply restoration control

mode ............................................................................................................................... 41

Figure 3.4: Two-feeder MV distribution network with a back-to-back VSC based SOP

........................................................................................................................................ 42

Figure 3.5: Transient response of the power flow control mode to step changes in active

and reactive power references: (a) DC side voltage; (b) reactive power response of

VSC1; (c) active power response of VSC2; (d) reactive power response of VSC2 ...... 44

Figure 3.6: SOP response for a three-phase fault at t=1s and blocked at t=1.2s: (a)

output voltage (left) and current (right) on the faulted side VSC; (b) output voltage (left)

and current (right) on the un-faulted side VSC; (c) power flow behaviour on both VSCs

........................................................................................................................................ 47

Figure 3.7: SOP response for a single-phase to ground fault at t=1s and blocked at

xii

t=1.2s: (a) output voltage (left) and current (right) on the faulted side VSC; (b) output

voltage (left) and current (right) on the un-faulted side VSC; (c) power flow behaviour

on both VSCs ................................................................................................................. 49

Figure 3.8: Control mode transition system ................................................................... 50

Figure 3.9: Results of a hard transition to supply restoration mode at t=2.2s: (a) output

voltage waveform of the faulted side VSC; (b) output current waveform ..................... 51

Figure 3.10: Output current waveforms of the faulted side VSC for a hard transition to

supply restoration mode with: (a) an inductance connecting to Bus 17; (b) a capacitance

connecting to Bus 17 ...................................................................................................... 52

Figure 3.11: Results of a smooth transition to supply restoration mode at t=2.2s: (a)

output voltage waveform of the faulted side VSC; (b) output current waveform ......... 53

Figure 3.12: Results of a hard transition to power flow control mode at t=3s: (a) output


Figure 3.13: Synchronisation controller ......................................................................... 55

Figure 3.14: Voltage synchronisations for reconnection: (a) phase synchronisation; (b)

magnitude synchronisation............................................................................................. 56

Figure 3.15: Results of a smooth transition to power flow control mode: (a) output


Figure 4.1: (a) Simple distribution network with an SOP; (b) Power injection model of

SOP for distribution network power flow control .......................................................... 61

Figure 4.2: Comparison between the quadratic and approximate loss estimation

function .......................................................................................................................... 64

Figure 4.3: Flow chart of the proposed PDS method ..................................................... 68

Figure 4.4: Example of one SOP optimization by PDS method .................................... 71

Figure 4.5: Flowchart of proposed network reconfiguration considering SOPs ............ 73

Figure 4.6: 33-bus distribution network [99] ................................................................. 74

Figure 4.7: Impact of different number of SOP installation on power loss minimization

and voltage profile improvement ................................................................................... 76

Figure 4.8: Impact of different number of SOP installation on load balancing and

xiii

voltage profile improvement .......................................................................................... 77

Figure 4.9: Voltage profiles of the network under normal loading condition ................ 80

Figure 4.10: Results of load balancing capability and relevant voltage profile

improvement under different methods ........................................................................... 82

Figure 4.11: Voltage profiles of the network under different cases with DG connection

........................................................................................................................................ 83


improvement under different cases with DG connections ............................................. 85

Figure 4.13: Impacts of the SOP device losses on total network power losses ............. 87

Figure 5.1: Voltage profiles of an MV feeder with and without DG.............................. 91

Figure 5.2: An MV feeder with segments for the minimum voltage estimation ............ 92

Figure 5.3: A part of distribution feeder ......................................................................... 93

Figure 5.4: Voltage estimation using simplified method ................................................ 95

Figure 5.5: Distributed voltage control framework........................................................ 96

Figure 5.6: Interior structure of a local control agent .................................................... 97

Figure 5.7: Details of local measurements ..................................................................... 98

Figure 5.8: Radial distribution feeder ............................................................................ 99

Figure 5.9: Two-feeder MV distribution network ........................................................ 101

Figure 5.10: Voltage profiles of both feeder before voltage control ............................ 102

Figure 5.11: Voltage profiles with SOP voltage control: (a) active power control method;

(b) reactive power control method ............................................................................... 103

Figure 5.12: Comparison of the network active power losses ..................................... 104

Figure 5.13: Priority-based coordination of the multiple local control agents ............ 105

Figure 5.14: Prioritisation of the proposed objectives from high (H) to low (L)......... 106

Figure 5.15: Flow chart of the proposed priority-based coordination ......................... 107

Figure 5.16: The four-feeder MV test network ............................................................ 118

Figure 5.17: Normalized profiles of a weekend and weekday: (a) loads; (b) DGs ...... 118

Figure 5.18: Per unit maximum and minimum voltage profile over the studied period at

(a) feeder 1; (b) feeder 2; (c) feeder 3; (d) feeder 4 ..................................................... 120

xiv

Figure 5.19: Per unit voltage profile at 11am in the first studied day along (a) feeder 1;

(2) feeder 2; (3) feeder 3; (4) feeder 4 .......................................................................... 121

Figure 5.20: maximum and minimum voltage profiles at different control schemes .. 122

Figure 5.21: Change of the tap operation at different control schemes ....................... 123

Figure 5.22: Total network losses at different control schemes ................................... 124

Figure 5.23: Apparent power requirements for different SOP control schemes .......... 125

Figure 5.24: maximum and minimum voltage profiles when increase DG penetration

...................................................................................................................................... 126

Figure 5.25: maximum and minimum voltage profiles with increased DG penetration

by coordinated control with and without involving SOP ............................................. 127

Figure 5.26: Total active power generated with increased DG penetration by

coordinated control with and without involving SOP .................................................. 127

Figure 5.27: Change of the tap operation with increased DG penetration at different

cases. ............................................................................................................................ 128

Figure 5.28: Total network losses with increased DG penetration .............................. 128

Figure 5.29: Power settings with increased DG penetration: (a) apparent power settings

of the SOP, (b) reactive power settings of the PV DG unit .......................................... 129

xv

List of Tables

Table 3.1: Parameters of the Back-to-back VSC Based SOP ........................................ 42

Table 4.1: System Load Balancing Index with Different Number of SOP Installation . 76

Table 4.2: Results of Computing Time for Optimal SOP Operation ............................. 78

Table 4.3: Results of Different Methods for System Power Loss .................................. 79

Table 4.4: Results of Different Methods for Load Balancing ........................................ 81

Table 4.5: Results of Different Methods for Power Loss Minimization with DG

Connections .................................................................................................................... 83

Table 4.6: Results of Different Methods for Load Balancing with DG Connections .... 84

Table 5.1: Data of the Load and DG ............................................................................ 118

Table A.1 Network Parameters for the SOP study in Chapter 5 .................................. 149

Table B.1 Network Parameters for the Case study in Chapter 5 .................................. 151

xvi

Nomenclature

List of abbreviations

AC Alternative Current

CI Customer Interruptions

CML Customer Minutes Lost

DC Direct Current

DG Distributed Generation

DMS Distribution Management System

DNO Distribution Network Operator

DVR Dynamic Voltage Restorer

d-q Direct-Quadrature

EV Electric Vehicles

FACTS Flexible AC Transmission Systems

GA Genetic Algorithm

GHG Greenhouse Gas

HVDC High-voltage DC Transmission System

ICT Information and Communication Technologies

IGBT Insulated Gate Bipolar Transistor

LBI Load Balance Index

MPPT Maximum Power Point Tracking

MV Medium Voltage

NOP Normally Open Point

xvii

OFGEM Office for Gas and Electricity Market

OLTC On-Load Tap Changer

PDS Powell’s Direct Set

PI Proportional Integral

PLL Phase Locked Loop

PV Photovoltaics

PWM Pulse Width Modulation

SCADA Supervisory Control and Data Acquisition

SOP Soft Open Point

STATCOM Static synchronous compensator

UPFC Unified Power Flow Controller

VSC Voltage Source Converter

Chapter 1 Introduction

1

Introduction

HIS chapter introduces the background, motivation, objectives, and

contributions of this thesis. An outline of the thesis is also provided.

Chapter 1

T


2

1.1 Background

Electric power systems, after several years of slow development, are experiencing

tremendous changes. In response to environmental and climate change, renewable

generation is expected to play an important role in future power generation systems.

The continuous development of economy as well as large-scale integration of

electrified heating and transport are leading to increased demand for electricity. To

tackle the inadequacies of current electricity grid, more intelligence is required in the

grid for the sake of security, economy and efficiency thus allowing the emergence of

the smart grid concept.

1.1.1 Climate Change and Renewable Generation

Global climate change resulting from greenhouse gas (GHG) emissions has had

observable effects on the environment, such as loss of sea ice, frequent wildfires and

more intense heat waves [1].

Over the past decades, worldwide efforts have been made in preventing the climate

change. In 1997, more than 160 countries were gathered in Kyoto to discuss the

solutions and signed an agreement, named ”Kyoto Protocol”, in order to reduce the

emissions of GHG (carbon dioxide (CO2), methane (CH4), nitrous oxide (N2O)) to at

least 5% below 1990 level during the commitment period of 2008 -2012 [2]. In 2012, a

new deal to extend the Kyoto Protocol was agreed by nearly 200 countries, continuing

cutting the GHG emissions during the 2013- 2020 period. In 2009, the European

Commission ratified an agreement, named renewable Energy Directive, targeting a 20 %

reduction of GHG emissions by 2020 compared with 1990 level for the European

Union (EU) member countries. In addition, a long-term goal was established to cut its

emissions substantially by 80-95% compared to 1990 by the end of 2050 [3].

The UK government has also committed to the reduction of GHG emissions targeting


3

at both international and EU level. The government has targeted a reduction of CO2

emissions of 34% (below 1990 level) by 2020 and at least 80% by 2050. In order to

meet these targets, a number of steps have been taken. The carbon budgets are used to

make sure the UK is on track [4]. The impact of these regulatory schemes has been

enhancing the use of renewable energy and the ‘clean’ generation technologies. Figure

1.1 presents an illustrative breakdown of the final shares of different types of renewable

sources and technologies in the UK in 2020 [5].

Figure 1.1: The UK Renewable Energy Target [5]

The overall targeted increase in the amount of energy generated from renewable

sources will be from 2% to 15% by 2020. This target, as suggested in [5], could be

achieved by the following renewable energy targets in each energy consumption sector:

more than 30% of electricity generated from renewables;

12% of heat generated from renewables;

10% transport energy from renewables.

To assist the delivery of UK’s ambitious targets, a substantial increase in using

renewable energy, in the form of distributed generators (DGs), has emerged in the

electricity sectors. DGs are small-size electricity power generation connected to a

distribution network rather than the transmission network [6]. They have lots of

economic and technique benefits for electric power networks, such as reducing network


4

losses, improving reliability and deferring network infrastructure reinforcement. [7].

Nevertheless, from the network operation perspective, problems are introduced by

connecting a large capacity of DGs to the network, such as the reverse power flow,

violation of voltage and thermal limits, increase of fault levels, harmonics and flickers

etc.

1.1.2 Electricity Demand Growth

Electricity demand is the fastest-growing component of total global energy demand due

to the increase of population, development of economy and the fact that electrical

energy is the most suitable form of the energy with respect to human activities and

environment [8]. As shown in Figure 1.2, the global demand of electricity increased by

over 50% from 1990 to 2011, and is expected to increase by over 80% from 2011 to

2030 under current policies scenario. Among all the continents, Asia, as a developing

area, has the most radical increase in electricity demand, projected to average around 5%

per year respectively to 2035 [9].

Figure 1.2: World Electricity Consumption by Region [9]

The UK’s demand for electricity is expected to increase over the next a few decades

[10]. A key reason is the transition to low carbon future. To meet government carbon


5

reduction targets, the UK’s dependence on fossil fuels (e.g., gas and petrol) would be

reduced and changed to renewable sources (e.g., wind and solar power). Examples

include the large-scale integration of electrified transport (e.g., electric vehicles) and

decarbonized domestic heating (e.g., electric heating pumps). Because of these, the

UK’s demand for electricity could double by 2050 [10].

To meet the growing demand for electricity, additional network capacity is needed. This

would mean significant investments if using the conventional network reinforcement

solution. Two direct consequences are anticipated [11]:

High cost to customers

According to the estimation made by Office for Gas and Electricity Market

(Ofgem) [12]: the UK needs to invest as much as £53.4 billion in the electricity

network between 2009 and 2025. This investment would be paid eventually by

customers.

Impact on society and the environment

Reinforcing the electricity network will have a significant impact on carbon

emissions [10]. Additionally, the reinforcement usually involves extensive

excavations and disruption, while significant amount of time is inevitable due to

the work scale (e.g. cable upgrades).

Under this circumstance, innovative solutions should be sought without or with reduced

network reinforcement and expansion.

1.1.3 Emergence of Smart Grid Concept

Utility engineers across the globe are trying to address numerous challenges, e.g. the

increasing integration of renewable generation, the continuing growth of electricity

demand and the need for optimal deployment of expensive assets. It is evident that


6

these challenges cannot be solved based on the existing electricity grid. New generation

of electricity grid, known as “Smart Grid”, is being developed to address the major

shortcomings of the existing electricity grid [13].

According to the US Department of Energy’s Modern Grid Initiative:

“A smart grid is an electricity network that employs innovative products and service

technologies together with advanced monitoring, control and communication

technologies in order to:

Enable and motivate active participation by consumers;

Accommodate all generation and energy storage options;

Enable new products, services, and markets;

Run more efficiently;

Provide higher quality of power required;

Anticipate and respond to system disturbances in a self-healing manner;

Operate resiliently against attack and natural disaster.” [14, 15]

The UK’s Electricity Networks Strategy Group (ENSG) published “A Smart Grid

Vision” to outline what a UK smart grid should be and what challenges need to

addressed. According to it, a Great Britain smart grid should be able to help the UK

meet its carbon reduction targets, ensure energy security and wider energy goals while

minimising costs to consumers. A “Smart Grid Routemap” was also published to

deliver this vision [16]. It emphasizes the potential role of the smart grid as a key

enabler to help UK power system accommodate a number of critical developments in

the next 40 years. The detailed developments with an illustrative timeline are given in

Figure 1.3 [17].


7

Figure1.3: Critical developments for UK smart grid routemap out to 2050 [17]

In the UK, Smart Grid has been primarily focused on the distribution networks where it

is believed early action is needed. Firstly, the distribution network is the biggest

component of electricity losses. It is most important that the distribution network

operators (DNOs) can manage their carbon footprint. Moreover, the uptake of electric

heat pumps and electric vehicles, investment in small wind generation, household and

community micro-generation and other initiative to de-carbonize the energy use could

have a profound impact on the nature and pattern of demand on the distribution

networks. DNOs may need to actively manage the intermittent and bi-directional power

flows of electricity, which requires significant changes in electricity generation and

demand technologies [16]. Therefore, initially, the distribution network will have the

greatest opportunity for smart interventions where mass investment is required, aligned

with revised network operation and control paradigms, to ensure it can cope with the

increasing demands and many new emerging requirements [16].

1.2 Research Motivation

Medium Voltage (MV) distribution networks are usually operated in a radial

configuration. Normally Open Points (NOPs) are built, connecting adjacent feeders, to

provide alternative routes of electricity supply in case of planned or unplanned power


8

outages. The unique benefit of this configuration is its inherent simplicity of operation

and protection. However, this needs to be reconsidered under the new circumstance.

With increasing penetration of intermittent renewables and growing electricity demand,

distribution networks are facing the challenge of increasing power flows through

existing networks. Although traditional network reinforcement is an option to deliver

the additional network capacity, their high-cost and time-consuming features make this

solution undesirable. Innovative solutions are therefore needed to increase the

utilization of existing assets dynamically and reroute the power flow through less

loaded circuits.

One way is to interconnect (or mesh) the conventional radial network, which enables

the power delivery through a less heavily loaded feeder and hence relieves the stress on

the heavily loaded feeders. Currently this configuration has been recognized by several

DNOs in their development plans [18], achieved by closing the NOPs. Benefits of a

closed loop network include balancing the loads between different feeders, improving

the voltage profiles, reducing power losses and improving the reliability. However, the

main problems of interconnection are:

The rise in fault current, which could lead to the malfunction of the existing

circuit breakers.

The need of more complicated and expensive protection schemes.

An alternative solution between the radial and the mesh operation is to have a flexible

mesh configuration. It is achieved by using a power electronic based device, named

Soft Open Point (SOP), to replace the NOP. Instead of simply opening/closing the NOP,

an SOP is able to provide the following functionalities:

Flexible control of active power exchange between connected feeders;

Flexible manipulation (absorbing and supplying) of reactive power on both

interface terminals;


9

Immediate fault isolation between interconnected feeders;

Small and controllable contributions to fault current;

Immediate post-fault supply restoration to support isolated loads on a feeder

through power transfer from the adjacent feeder.

Therefore, the SOP offers further flexibility to current distribution networks. For

network operation, it is envisioned to release capacity of existing feeders, enhance the

efficiency and service quality to customers. This thesis is motivated to investigate the

following three points, regarding the use of SOPs for MV distribution network

operation.

1.2.1 Control of an SOP

The SOP is a multi-functional power electronic device, and hence its control strategy

should be able to realize all the aforementioned functionalities. Previous studies (the

detailed discussion is in Section 2.3) mainly focused on the use of SOP under normal

network operating conditions, where the current-controlled strategy was used to

achieve independent control of real and reactive power. However, this

current-controlled strategy is not suitable to provide post-fault supply restoration

(another functionality of the SOP) because it may result in voltage and/or frequency

excursions that lead to either unacceptable operating conditions or instability. The

operating principle of an SOP therefore needs to be investigated under both normal and

abnormal network operating conditions.

1.2.2 Benefits analysis of using SOPs

It is urgently needed to quantify the benefits of SOPs for distribution network operation,

which provides confidence for using SOPs in distribution networks. SOPs are effective

in reducing power losses, balancing feeder loads, improving voltage profile, and

thereby increasing network loadability and DG connections. A number of studies (the


10

detailed discussion is in Section 2.3) can be found for these purposes. However, they

are mainly based on installing one or two SOP in a simple two-feeder network together

with the controller design and simulation. Methodologies for the benefit quantification,

i.e., steady state analysis of distribution networks with SOPs were not addressed and

the advantages of a more widespread use of these devices in distribution networks have

not been explored. In addition, distribution network reconfiguration has been

intensively researched. It could achieve similar objectives as the SOP. However, unlike

the SOP control, network reconfiguration is achieved by changing the network

topology: closing some NOPs while opening the same number of closed switches to

maintain a radial network structure. Extensive research has been conducted into

network reconfiguration (the detailed discussion is in Section 2.2.1) for feeder load

balancing, loss minimization and voltage profile improvement. It is needed to have a

comparison between using SOPs and network reconfiguration to enhance the operation

of a distribution network.

1.2.3 Distribution network voltage control with SOPs

Increasing DG capacity in the distribution network could cause the violation of voltage

limits. SOPs are able to assist voltage control through changing the real and reactive

power flow. Previous studies (detailed discussion is in Section 2.3) have reported the

potential of using SOPs for network voltage control. However, the voltage control

strategy considering the detailed operation of a SOP (determining the set points of the

SOP) has not been investigated. In addition, coordination with other network

controllable devices could achieve a better performance. There is a clear need to

develop a coordinated voltage control strategy, in which the SOP and other voltage

control devices are used together to achieve a better network voltage control.


11

1.3 Objectives and Contributions of this Thesis

The objectives and contributions of this work are outlined as:

Investigate the operating principle of an SOP using back-to-back VSCs under

both normal and abnormal network operating conditions.

Two control modes were developed for the operation of a back-to-back VSC

based SOP. The operating principle of the back-to-back VSC based SOP was

investigated under both normal and abnormal network operating conditions. The

performance of the SOP using two control modes has been analysed under

normal conditions, during a fault and post-fault supply restoration conditions.

Investigate the benefits of a MV distribution network with SOPs under normal

network operating conditions, focusing on feeder load balancing, power loss

minimization, and voltage profile improvement.

A steady state analysis framework was developed to quantify the benefits of

SOPs in MV distribution networks. The framework considers the network

reconfiguration and the combination of both SOP and network reconfiguration.

Investigate the use of the SOP for network voltage control and its coordination

with OLTC and DG units in an MV distribution network. The coordinated

voltage control strategy is able to maintain the network voltage while mitigating

DG active power curtailment, reducing the tap operations and the network

power losses.

A novel coordinated voltage control strategy was proposed considering the SOP,

the OLTC and multiple DG units. A distributed control framework was used,

considering each of the controllable devices as a local control agent. A


12

priority-based control was developed, achieving the specified objectives in

accordance with their prioritisation. Unlike centralised control strategies

(detailed discussion in Section 2.2.2), which requires communication links with

each network node and power flow solutions at each time step, this coordinated

voltage control strategy achieves a trade-off among objectives with fewer

communicational requirements and less computation burden.

1.4 Thesis Outline

The rest of this thesis is organized as follows:

Chapter two provides a literature review of challenges for distribution network

operation. Available network operation functions in facing the emerging challenges are

then introduced, with special attention paid to the network reconfiguration, coordinated

voltage control and those functions offered by using power electronic devices in

distribution networks. Last, the state-of-the-art of SOPs, including the benefits, possible

device types and previous studies are presented.

Chapter three describes two control modes for the operation of a back-to-back

VSC based SOP. The operating principle and performance of the back-to-back VSC

based SOP using these two control modes were analysed under both normal and

abnormal network operating conditions.

Chapter four describes a steady-state analysis framework to quantify the

operational benefits of a distribution network with SOPs. A generic model of an SOP

for steady-state analysis was developed. Based upon it, an improved Powell’s direct set

method was developed to obtain the optimal SOP operation. Distribution network

reconfiguration algorithms, with and without SOPs, were developed and used to

identify the benefits of using SOPs. Case studies have been conducted with promising


13

results obtained for using SOPs in MV distribution networks.

Chapter five describes a novel coordinated voltage control strategy, considering

SOPs, OLTC and DG units, to maintain network voltages while mitigating DG active

power curtailment, reducing the tap operations and the network power losses. A

distributed based voltage control framework was developed which uses a simplified

voltage profile estimation method. Followed by an analysis of using SOP for

distribution network voltage control, a priority-based coordination, using the proposed

voltage control framework, was developed to achieve a trade-off among the objectives.

The effectiveness of the proposed coordinated control strategy was verified under

various loading and DG penetration conditions.

Chapter six presents the conclusions drawn, main findings and recommendations

for future work.

Chapter 2 Literature Review

14

Literature Review

HIS chapter presents a literature review on the challenges for

distribution network operation, the network operation functions and

the state-of-the-art of the SOP studies.

T

Chapter 2


15

2.1 Challenges for Distribution Network Operation

Since the Carbon Emission Reduction Target (CERT) was introduced into the electric

industry, electric power distribution networks have undergone dramatic changes. In this

CERT context, distribution networks should not only be able to deliver electricity to

customers with acceptable reliability and quality standards, but also to fulfil the tasks, such

as accommodating increasing renewable energy sources and meeting the sustained high

level of demand growth resulting from the electrified heating and transport, etc. These

requirements impose great challenges to distribution networks, in particular in the

operation perspective.

2.1.1 Aging Assets and Lack of Circuit Capacity

In many parts of the world (e.g., the USA and most European countries), the electric

power system expanded rapidly in 1950s and 1960s [19]. The distribution equipment

and cables installed then are now beyond its design life and in need of replacement. The

capital costs of the like-for-like replacements will be very high [19]. The need to

refurbish the transmission and distribution circuits is an obvious opportunity to

innovate with new designs and operating practices. In addition, the introduction of

intermittent DGs can increase the use of tap-changing devices (e.g., OLTC and

switched capacitor banks) and therefore accelerate failure of the device [20]. A smarter

operation is therefore needed to make effective use of existing assets and mitigate the

wear and tear.

In many countries, constructing new overhead lines, needed to meet the load growth

and to connect renewable DGs, has been delayed for up to 10 years due to the

difficulties in obtaining rights-of-way and environmental permissions [19]. Therefore,

some of the existing distribution line circuits are operating near their capacity, limiting

DG integration and demand growth. This calls for intelligent methods, instead of

traditional network reinforcement, to increase the power transfer capacity of circuits


16

dynamically and reroute the power flows through less loaded circuits.

2.1.2 Operational Constraints

Distribution networks have to operate within prescribed constraints in order to ensure

secure electricity delivery. These constraints include [21]: voltage constraints, thermal

constraints, frequency and short-circuit constraints, etc.

With the introduction of DGs and the continuous growth of demand, most distribution

networks are experiencing constraint breaches. The large-scale integration of DGs can

cause over-voltages at times of light load, and the demand growth can give rise to local

low voltages. Hence, adequate voltage management requires a coordinated operation of

the DGs, on-load tap changers and other equipment (e.g., voltage regulators and

capacitor banks). Due to the local reverse power flow brought by the DGs, aligned with

the increased demand, thermal limits of the existing transmission and distribution

equipment and lines can be exceeded. Thus dynamic ratings that can increase circuit

capacity in real-time are needed to address the thermal overloading. Some DG plants

use directly connected rotating machines, for example, old wind turbines and small

hydro turbines use induction generators while, large hydro turbines use synchronous

generators [22]. The connection of these generators and new spinning loads can cause

the maximum fault current to exceed the fault level rating of the distribution network.

Thus effective fault current limiting technologies or services are required to maintain

safe operation of the power system.

Therefore, distribution network operation has to be robust and flexible to address the

constraint problems that are caused by the quantity, geography and intermittency of the

DG integration as well as the new forms of load. Otherwise, traditional expensive

network reinforcement has to be carried out.


17

2.1.3 Reliability and Efficiency of Supply

Morden society has increasing critical loads connected and power blackouts could

result in enormous monetary loss and society chaos [23]. In the coming decades, the

proportion of intermittent renewable generation in the energy mix will get higher,

lowering the overall predictability of the electricity supply. In the meantime, the

electricity capacity margin is reducing, which increases the risks to security of supply.

To meet the requirement of the Government’s reliability standard, UK’s utilities had

defined customer interruptions (CIs) and customer minutes lost (CMLs) as the indicator

of each DNO’s performance [24]. This calls for effective restoration strategies and

other post-fault measures to retain or return the supply after the (inevitable) faults in the

distribution networks.

Electrical power losses are an inevitable consequence of transferring electricity across

distribution networks and they have significant environmental and financial impact on

customers. For instance, the energy losses from the electricity distribution networks

contribute to approximately 1.5 % of Great Britain’s greenhouse gas emissions [25].

Moreover, present distribution networks usually have power losses in the range of 3% -

9% and this accounts for 80% of the total power losses in combined transmission and

distribution networks [24]. Higher losses (dominated by the ohmic loss: I2R terms)

would be expected in 2030 due to the anticipated more heavily assets use [24]. All

these losses are eventually paid for by customers and need to be mitigated. To

encourage an efficient level of losses in distribution networks, Office for gas and

electricity Market (OFGEM) has launched a losses incentive mechanism to be a part of

their network price control since 2015 [26]. An improved operating efficiency is

therefore needed for minimum energy losses.


18

2.2 Network Operation Functions

As discussed in the previous section, distribution networks need better control and

supervision of distribution network facilities to support their operation in facing the

emerging challenges.

Traditionally, there are well established operation functions available that enable

control and supervision of distribution network facilities for desired network

performance. Examples include the network reconfiguration, voltage and reactive

power control, protection relay coordination and outage management, etc.

At present, the emergence of smart grid concept is a radical reappraisal of the functions

of distribution networks. In response to the smart grid initiatives, various new

technologies and applications are being introduced into distribution networks, such as

the integration of distributed energy resources, active control of demand load and the

progress of using distribution-level power electronics. These new technologies and

applications are able to support network operation neither by adding to the

effectiveness of existing operation functions (e.g., through a coordinated operation) or

by offering new functions to the distribution network.

This thesis focused on the enhancement of distribution network operation through the

use of power electronic based equipment - SOP - as well as its operation comparing and

coordinating with traditional network reconfiguration and network voltage control. As

background to the research that has been carried out, network reconfigurations,

coordinated voltage control and the operation functions that can be offered by using

power electronic devices in distribution networks will be introduced in subsequence

sections.


19

2.2.1 Network Reconfiguration

Distribution networks are normally designed as meshed networks but are operated in a

radial manner with normally open points [27]. Network configurations can be changed

during the operation by changing the open/close status of switchgears, manually or

automatically [27]. Figure 2.1 gives an example of network reconfiguration, which is

based on a three-feeder network with three NOPs (dash lines). The configuration of the

network is changed by closing NOPs 15 and 26 while opening closed switches 12 and

13.

(a) (b)

Figure 2.1: Three-feeder example network: (a) before network reconfiguration; (b)

after network reconfiguration [27]

The main objectives of network reconfiguration include [19, 28]:

1. Supply restoration：This optimally restores de-energized customers through

alternative sources in the case of planned and unplanned power outages.

2. Load balancing among substation transformers or different feeders and

equalising the voltages.

3. Active power loss minimization at a given time or energy loss minimization

during a period.


20

The methods used to solve the above network reconfiguration problems include those

based on practical experience and optimization techniques. The methods based on

optimization techniques determine the optimal network configuration by solving a

complicated combinatorial, non-differentiable constrained optimization problem [29].

Possible solutions include mathematical algorithms [27, 30], computational

intelligent-based algorithms (for example, generic algorithm, simulated-annealing,

fuzzy logic) [31-33], and hybrid algorithms which combine two or more above

algorithms [34, 35].

Network reconfiguration and its significant benefits have been reported in the past

decades. However, practical applications of automatic network reconfiguration is

presently very limited due to excessive costs of remotely-controlled switchgear,

associated ICT infrastructures and maintenance of hardware/software (e.g., excessive

wear and tear of switches) [36, 37]. The protection coordination requirement for the

reconfigurable network is also a hurdle [38].

It is possible to use SOPs to achieve the same aforementioned benefits (objectives)

offered by network reconfiguration. In the meantime, changes of network topology and

protection coordination are no longer needed when using the SOP. In the study reported

in Chapter 4, the potential and benefits of using SOP, compared with traditional

network reconfiguration, was discussed in further detail.

2.2.2 Coordinated Voltage Control

Voltage control refers to the technique of using available voltage control equipment to

maintain acceptable voltage levels at all points in the distribution network under all

loading conditions [39]. As discussed in Section 2.1, it is one of the most significant

issues that limits DG penetration in distribution networks. When DGs units are

connected to a distribution network, they can significantly change the network voltage

profile and interfere with conventional local control strategies of the transformer


21

tap-changers, line voltage regulators and shunt capacitors (which are designed based on

the assumption of unidirectional power flow [40]). This interference leads to [41]: 1)

unexpected over- and under-voltage, 2) increase in power losses and 3) excessive wear

and tear of voltage control devices.

In these circumstances, a coordinated operation of the voltage control devices is needed

to provide adequate voltage management, meanwhile, eliminating the aforementioned

problems. In the literature, numerous control devices has been conducted to support

voltage control. Examples include [42-45]: on-load and off-load transformer

tap-changers, line voltage regulators and switched capacitor banks, energy storage,

power electronic devices and the control of DG.

These control devices are the key factors to realize the coordinated voltage control in

distribution networks. However, the first milestone is to find a most effective control

strategy to achieve the coordination among the control devices. In the literature,

numerous voltage control strategies have been proposed to achieve a proper

coordination. According to the control structure and communication links, they can be

categorized as [46]:

i) Localized coordination: with no communication links,

ii) Centralized coordination: with a wide range of communication links,

iii) Distributed coordination: with a few communication links among control

devices.

Localized coordination

Localized coordination does not require communication links. Control devices

determine their operating set points based on local signals [39]. The coordination

between the localized control actions can be achieved based on the time delay operation

[40]. Alternatively, the local control strategies for DG units can be modified to mitigate


22

their impacts on the voltage profile, without modifying conventional local control

strategies. For example, the authors in [47] proposed a decentralized reactive power

control approach for the DG units, which absorbs reactive power to compensate the

effect of voltage rise caused by active power injection. Therefore, the convention local

control strategy is not affected by the reverser power flow.

The localized coordination approaches are reliable and scalable due to their

independency and simplicity of control structure. However, their control performances

are not optimal. Examples include: 1) the high stress on the control devices is not

relieved, 2) the energy capture from DG units is not maximized, and 3) the feeder

power losses may increase.

Centralised coordination

Centralized coordination is a fully-coordinated control approach that calculates the

optimal operating set points for all available controllable devices and delivers the best

performance possible by solving a multi-objective optimization problem [48]. It is

usually achieved based on a distribution management system (DMS) to receive

measurements of the distribution network (e.g., voltage, power flow, and equipment

status), and, then, to estimate the network status before dispatching control devices

using state estimation technologies [49].

Figure 2.2 shows an example of the cascaded control architecture for centralized

coordination. The distribution management system solves the optimization problem by

either following different objectives at different times or considering conflicting

objectives together in a weighted manner [50-54]. The DMS then dispatches the control

devices accordingly via the supervisory control and data acquisition (SCADA) systems.

The main downsides to the centralized coordination include [49, 55]: 1) The extensive

investment in sensors and communication links; 2) The requirement for the accuracy of


23

the state estimator; 3) The use of power flow and optimization solution at each time

step, which might cause a large computation burden and numerical convergence issues

when X/R ratio is low; 4) The undesirable properties with respect to scalability and

reliability (fragile to single point of failure).

Figure 2.2: Cascaded control architecture for centralized coordination [48]

Distributed coordination

Distributed coordination has been proposed recently to tackle the drawbacks of

centralized and localised coordination [55]. It is usually accomplished by using a

multi-agent framework. Each agent in the framework is an intelligent entity that can

perceive its environment, create an action according to its own decision-making and

communicate with other agents to achieve a common goal [46]. For voltage control

problems, each control device is considered as an local control agent and communicates

with other agents for improved control performance [46].

Compared with the centralized coordination, distributed coordination can reduce the

communication requirements and computation burden as well as avoid the reliability

issue (for example, single point of failure) because the dispatching is implemented in a


24

distributed manner, i.e., the original problem is decomposed into smaller sub-problems

and each is assigned to a particular agent [56]. In addition, the implementation of

centralized coordination becomes difficult with the increase of uncertainties due to DG

integration and the efforts to achieve a “plug and play” property. Distributed

coordination is expected to deal or at least relieve these issues [56]. In this thesis, the

distributed coordination was employed to provide coordinated voltage control in an

active distribution network with SOPs.

Numerous efforts have previously been made to achieve distributed coordination for

network voltage control using the multi-agent framework [41, 46, 55-57]. In [56], a

multi-agent reactive power dispatching scheme was proposed for voltage support of

DG units in a single feeder. But the utility voltage control devices (for example, OLTC,

shunt capacitors) were not considered. The authors in [57] proposed a coordinated

voltage control for the OLTC control, where the remote terminal units were treated as

local control agents and provide distributed state estimation. However, the OLTC was

assumed as the only control devices. In [41] and [55], multi-agent frameworks were

proposed for coordinated voltage control between both DG units and utility control

devices. Figure 2.3 shows an example of the coordination among three control agents

via two ways communication [41]. All three control agents have a common objective

and each agent also has its own objective (see the circles). The control agent can take

local measurements, make control decisions and negotiate with other agents when there

is a need. In these studies, the control performances highly depend on their control

design since the control optimality tends to be degenerated by the negotiation process

among agents.

Another coordinated control strategy using a multi-agent framework was proposed in

[46] to achieve both autonomy and optimal control. The use of “blackboard” memory

data combined with the multi-agent framework can eliminate unnecessary negotiation

process among agents and minimize data communications. However, the extensive


25

investment on sensors, the computation burden and the convergence issues due to

running power flow were not addressed in this study.

Figure 2.3: Coordination among three control agents via two ways communication [41]

In this thesis reported in Chapter 5, based on the proposed distributed coordination, a

simplified voltage estimation method was used to reduce the number of sensors in the

network. In addition, a coordination method was applied on the control agents (SOP,

OLTC and DGs), which can eliminate unnecessary negotiation process among the

agents and provide improved control performance without running power flow.

2.2.3 Operation Functions Offered by Power Electronic devices

Power electronics, in forms of the high-voltage DC transmission system (HVDC) and

flexible AC transmission systems (FACTS), has played a visible and key role in

power-gird control for over 60 years, mainly on the transmission network for bulk

power transfer [58]. The use of power electronics at the distribution network level has

been much more limited. At present, under the smart grid concept, visibility,

controllability, and flexibility will be essential features throughout the future

distribution network where power electronics can play a key role [19]. Several DNOs

in the UK have already initiated to offer power electronics solutions to address their


26

network issues [59-61] and, hence, delay or eliminate the need of traditional network

reinforcement. Further literature on the advantages of power electronic solutions over

traditional network reinforcement is available in [18].

A wide variety of network operation functions can be offered by using power electronic

devices in distribution networks. Examples include voltage control, phase rebalancing,

active power filtering, fault current limiting and power flow control, etc.

Power electronics can support distribution network voltage control in forms of power

electronic transformers [62], transformers with electronic tap changers [63] and

reactive power compensation. Power electronic transformers and transformers with

electronic tap changers can be installed in place of classical transformers, providing the

competitive advantages of enabling fast and smooth voltage regulation as well as little

wear and tear. The reactive power compensation, such as using the STATCOM [64],

dynamic voltage restorer (DVR) [65] and SOP [61], can tackle the voltage problem

either at the substation, path-way along the feeder or at the point-of-load.

Phase imbalance, as a result of differences in consumption patterns of consumers, can

cause higher conduction losses and lower network utilization. It is also considered as a

major cause of capacity limitation [66]. Power electronic devices that contain a

common DG bus, such as the unified power quality conditioner, DVR, STATCOM and

SOP, allow exchange of instantaneous power between the phases and, hence, can be

used to rebalance current flow on a feeder.

Active power filters for harmonic elimination and power quality improvement have

been applied within customer premises but not as network elements compensating a

group of consumers [18]. In recent years, the need for filtering in networks with power

electronic compensation has been recognized by DNOs, such as filtering in the

locations with a high penetration of PV [18].


27

Power flow control are traditionally achieved through network reconfiguration with

existing switchgear (discussed in Section 2.2.1), aiming to increase the power transfer

capacity of circuits dynamically and reroute the power flows through less loaded

circuits. At present, an alternative solution, that makes use of power electronics in the

form of SOP to facilitate power flow in a dynamic and continuous manner, is attracting

the interest of the DNOs. A Low Carbon Networks Fund (LCFN) project has already

initiated to install the SOP in UK’s distribution network to verify its effectiveness [61].

This thesis focuses on the new power electronics application. Details on the

introduction of SOP benefits, types and previous work are presented in the following

section.

2.3 Soft Open Points

2.3.1 Benefits of Soft Open Points

As shown in Figure 2.4, SOPs are multi-functional power electronic devices (discussed

in Section 1.2) installed in place of the NOPs in distribution networks. They are also

referred to as ‘loop power flow controllers’, ‘series controllers’ and ‘DC links’ in some

literatures.

Figure 2.4: Simple distribution network: option A representing a NOP connection

and B representing an SOP [67]


28

The most significant advantageous provided by these power electronic devices,

compared with mechanical switches, are

The power flow between the connected feeders can be regulated in a continuous

and dynamic manner.

The reactive power is able to be independently injected or absorbed at both

interface terminals. Therefore, SOP can be used for dynamic voltage support.

The SOP can be used to improve the power quality of the distribution networks

[68], such as mitigating voltage imbalance, sags, flickers, and low-order

harmonics, etc.

The SOP can be used to connect any feeders regardless of the angular or rated

voltage difference between them, particularly when the feeders are supplied by

different substations. This is not possible with the mechanical switches.

The short-circuit currents on the radial feeders (without SOPs) are not changed

with the use of SOPs, owning to the capability of instantaneous fault current

control. This indicates that this intervention would not require modifications of

existing network protection assumptions or upgrades of protection devices (this

is usually the case for a close-loop operation). [69]

2.3.2 Types of Soft Open Points

The functionalities of an SOP were descripted in Section 1.2. Possible SOP types,

enable to realize these defining functionalities, include:

Back-to-back VSCs

Multi-terminal VSCs

Unified power flow controller (UPFC)

An overview of the topologies of these types of SOPs is shown in Figure 2.5. Each is


29

made from an arrangement of VSCs with varying rating and quantity. The VSC is

suitable for SOPs due to: 1) the freedom to operate with any combination of active and

reactive power, 2) the ability to limit (or control) fault current and 3) the possibility to

supply isolated areas of a network or even provide the black-start ability.

Figure 2.5: Topologies of different types of SOPs [70]

Back-to-back VSCs

This thesis focused on the back-to-back VSC based SOP. It consists of two VSCs

connecting via a common DC link to form an asynchronous AC/AC conversion device.

Coupling transformers are usually equipped, interfacing each VSC with the connected

feeder. However, transformerless topologies are possible [71], which reduces the size

and weight of the device. This arrangement allows active power exchange between the

AC front ends as well as independent reactive power manipulation at each interface

terminal.

Under abnormal network conditions, the fault on one feeder can be isolated from the

other feeder by the DC link. The fault current on the faulted side interfacing VSC can

also be limited (and controllable) by appropriate control strategy [68]. Post-fault

restoration is achievable, supplying the isolated loads on one feeder by power transfer

from the adjacent feeder via the VSCs [68].


30

Multi-terminal VSCs

The multi-terminal VSC based SOP is an extension of the back-to-back VSC based

SOP, in which additional VSCs are connected to a common DC-link. The number of

VSCs associated with a multi-terminal VSC based SOP is defined as 3 or greater. The

same features discussed for the back-to-back VSC based SOP are applicable to the

multi-terminal VSC based SOP.

Unified power flow controller (UPFC)

The UPFC based SOPs, also known as shunt-series VSCs, consists of two VSCs with a

common DC link. Unlike the configuration of the aforementioned two types, one VSC

in UPFC is connected in series and the other is connected in shunt with the

interconnected feeders. The series connected VSC injects a series voltage with

controllable magnitude and phase angle, thereby controlling the power flows between

interconnected feeders [72]. The shunt connected VSC supplies the active power

demanded by the series VSC through the DC link. It can also provide reactive power

manipulation independent of the operation of the series connected VSC.

The primary advantage of the UPFC based SOPs, compared with the aforementioned

two types, is the potential to induce power flows greater than VSC rating (for example,

10 MVA transferred using 1 MVA VSCs). However, this ability is determined not only

by the device ratings themselves, but also by the network topology, constraints, and

operating point as well as the device placement [70]. An inherent drawback to the

UPFC is the fact that the feeders connected with a UPFC are not isolated by a DC-link,

and, hence, disturbances or faults on one connected feeder will affect the other unless a

control strategy is developed to reject or compensate for disturbances. The

instantaneous fault current limitation (or blocking) is possible [73], at the cost of higher

device ratings. The interconnection is also not asynchronous and therefore the two

electrical network connected via a UPFC based SOP must be synchronous.


31

2.3.3 Previous Soft Open Points Studies

Previous work on SOPs is discussed with the research gaps identified in terms of the

control strategy, benefits analysis and their use for distribution network voltage control.

Control strategy

The control strategy of an SOP has been reported early in 2006 [71]. A

current-controlled strategy was developed for the operation of a back-to-back VSC

based SOP. In this control strategy, direct-quadrature (d-q) current components were

used for each VSC to provide control of instantaneous active and reactive power

exchange between the VSC and the AC network. The effectiveness of the control

strategy for instantaneous power flow regulation has been verified by a factory test

using 6.6kV 1MVA transformerless back-to-back VSCs. But the performance of this

control strategy under abnormal network conditions, i.e., during a fault and post-fault

supply restoration, was not investigated.

In [74], the same current-controlled strategy was applied to a multi-terminal VSCs

based SOP and validated in an experimental way. The transient performance of the

control strategy during a three-phase short-circuit fault was evaluated. This current

limiting capability has been demonstrated important to 1) protect VSCs from

over-currents and 2) limit fault current and therefore maintain the same short circuit

current as the radial network operation (without VSCs). Nevertheless, the operation of

the SOP for post-fault supply restoration was not considered.

When the SOP provides post-fault supply restoration to isolated loads, the frequency

and voltage are not dictated from the grid. Hence, the aforementioned control strategy

is not desirable because it may cause voltage and/or frequency excursions that can lead

to either unacceptable operating conditions or instability [75]. The operating principle

of an SOP under both normal and abnormal network operating conditions was not

found in previous literatures and need to be investigated.


32

Benefits analysis

The benefits of using SOPs in distribution networks have previously been reported for

different objectives, such as: loss minimization [76-78], feeder load balancing [36, 37]

and maximisation of network loadability and DG penetration [69, 79, 80]. The authors

in [69] developed an optimization framework to compare the benefits of using DC links

(which are also SOPs) with conventional network reinforcement measures. Using the

framework, a thorough assessment of the most promising SOP type was reported in

[79]. The benefits of the back-to-back VSC based SOP, UPFC based SOP and a series

VSC were analysed and compared in terms of their capability to maximise DG

penetration in distribution networks.

In these literatures, operational benefits of using SOPs were studied mainly based on

two or three feeders connected by an SOP. Methodologies for benefit quantification and

the advantages of more widespread use of SOPs in distribution networks have not been

explored.

On the other hand, it is possible to maximise the benefits of an SOP by combining with

traditional network reconfiguration (discussed in Section 2.2.1). An SOP is only able to

regulate the power flow of its own connected feeders. Using network reconfiguration

can change the feeders that are connected to the SOP dynamically and, hence, can

benefit the whole network. The combination of network reconfiguration and SOP

control and its benefits are discussed in further detail in Chapter 4.

Distribution network voltage control

The potential to use SOPs for distribution network voltage control has been reported

[67] [81] by changing active and reactive power flows on the connected feeders.

In [81], the minimum voltage deviation among multiple feeders was achieved by using

a multi-terminal VSCs based SOP. The effectiveness of using an SOP to mitigate


33

voltage deviation was compared with that using a close-loop operation. Similar

performance was observed while the later one requires more complicated protection

system. However, in this study, the reactive power control ability of the SOP was not

considered.

In [67], the use of SOPs, controlling both active and reactive power flow, has been

proposed for network voltage control and, ultimately, higher DG penetration in

distribution networks. A comparative analysis was conducted between using SOPs and

conventional voltage control options (for example, OLTC). However, the operation of

the SOP (set points of active and reactive power) was determined by merely satisfying

constraints. Well-established voltage control strategies including suitable operation

mechanisms of the SOP were not elaborated in this study. In addition, the coordination

of SOPs and other available voltage control devices for improved control performance

has not been investigated in previous work.

2.4 Summary

Challenges for distribution network operation were summarized. The network operation

functions that enable to deal with emerging challenges were then introduced, with

special attention paid to the network reconfiguration, coordinated voltage control and

those functions offered by using power electronics in distribution network level. The

power electronic applications can offer a variety of operation functions and, hence, will

play a key role to enable visibility, controllability, flexibility features throughout the

future distribution networks.

SOP is a power electronic application in distribution networks. The possible layout of

SOP and its benefits were introduced. In addition, a review of the state-of-the-art of the

SOP studies was presented and followed by identifying the corresponding research

gaps.


34

The current-controlled strategy has been reported for SOP operation to achieve the

defining functionalities. However, this control strategy is not suitable for post-fault

supply restoration (which is another key functionality of SOP). The operating principle

of an SOP under both normal and abnormal network operating conditions is yet to be

explored. Therefore, in this study, operating principle of a back-to-back VSC based

SOP was investigated under normal conditions, during a fault and post-fault supply

restoration (discussed in Chapter 3).

The benefits that an SOP can provide have been identified. However, methodologies for

benefit quantification and the advantages of more widespread use of SOPs in

distribution networks were not explored. In addition, using SOP control could achieve

the same objectives as traditional network reconfiguration, and a combination of both

SOP control and network reconfiguration could maximize the benefits of an SOP as

well as benefit the whole network. These issues have not been explored in previous

studies. Therefore, in this study, the operational benefits of using SOPs, with and

without considering network reconfiguration, were quantified and reported in Chapter

4.

Distribution network voltage control can be achieved by using SOP through both active

and reactive power flow control. However, research conducted on suitable operation

mechanism of SOPs for distribution network voltage control and the coordination with

existing voltage control devices was not found. Therefore, in this study, a coordinated

voltage control strategy in active MV distribution networks with SOPs was investigated

(discussed in Chapter 5), which is based on the distributed coordination.

Chapter 3 Control of a Back-to-Back VSC based SOP

35

Control of a Back-to-Back

VSC based SOP

HIS chapter investigates the operating principle of a back-to-back VSC

based SOP under both normal and abnormal network operating

conditions

Chapter 3

T


36

3.1 Introduction

In this chapter, the circuit topology of back-to-back VSCs and its suitability for the role

of an SOP is presented. Two control modes were developed for the operation of a

back-to-back VSC based SOP. The operating principle of the back-to-back VSC based

SOP with these two control modes was investigated under both normal and abnormal

network operating conditions.

Studies on a two-feeder MV distribution network showed the performance of the SOP

under different network operating conditions: normal, during a fault and post-fault

supply restoration. During the change of network operating conditions, a mode switch

method was used to achieve the transition between the control modes. The role and

importance of soft cold load pickup and voltage synchronisation processes, essential for

a smooth transition during the switching of control modes, were also investigated.

3.2 Back-to-Back VSC based SOP

Figure 3.1a shows a two-feeder distribution network interconnected with a

back-to-back VSC based SOP. Two VSCs are located between the feeder endpoints and

connected via a common DC link. The main circuit topology of the back-to-back VSCs

is shown in Figure 3.1b. It consists of a DC capacitor to provide an energy buffer and

reduce DC side voltage ripple, two two-level three-phase insulated gate bipolar

transistor (IGBT)-based VSCs to generate voltage waveforms using pulse width

modulation (PWM), and two series inductances [82]. These inductances provide

high-frequency harmonic attenuation, limit the rate of rise of short circuit current, and

facilitate control of power flow.


37

Feeder I

Real

power

Reactive

power

SO

P

Feeder II

MV distribution network

Reactive

power

(a)

+

-

Lf

VSC2

CLf

Lf

Lf

VSC1

Lf

Lf

(b)

Figure 3.1: (a) Basic configuration of a distribution network with an SOP; (b) Main

circuit topology of the back-to-back VSC based SOP.

The back-to-back VSCs are suitable for SOP operation due to the following

characteristics:

Flexible active and reactive power control

Both VSCs build their own voltage waveforms with desired amplitude and phase angle.

This allows a full control of active power flowing through the DC link as well as

independent reactive power supply or absorption at both interface terminals. This

controllability enables SOP operation for normal network operating conditions, which

includes feeder load balancing, power loss reduction and voltage profile improvement.


38

Instantaneous and independent voltage control

The voltage waveform built by the VSCs can be controlled dynamically within

milliseconds thus enabling transient control, e.g. dynamic Volt/VAR control and power

oscillation damping [83]. In addition, the VSC can build its own voltage without the

need of an active source at the receiving end. Hence cold load pick up for supply

restoration is achievable using such device.

Disturbance and fault isolation

Network disturbances and faults on one connected feeder can be isolated from the other

side by the back-to-back VSCs since transient overvoltage and overcurrent are able to

be limited by the control scheme [84].

3.3 Control Modes of the Back-to-Back VSC based

SOP

Two control modes were developed to operate the back-to-back VSC based SOP under

both normal and abnormal network operating conditions. The power flow control mode

was used to 1) regulate both active and reactive power flow on the connected feeders

under normal network operating conditions and 2) isolate fault between the

interconnected feeders when a fault occurs on one feeder. The supply restoration mode

was used under post-fault supply restoration conditions to provide power supply for the

isolated loads on one feeder through the other feeder.

3.3.1 Power Flow Control Mode

A dual closed-loop current-controlled strategy [85] was used to operate SOP for the

control of the feeder power flow under normal network operating conditions.

Current-controlled strategy is advantageous because: 1) it provides de-coupled control

of active and reactive power components; and 2) it inherently limits the VSC current


39

during network faults.

Figure 3.2 presents the overall control structure. The outer power control loop, the inner

current control loop and the PLL are the three main components. In the outer power

control loop (Figure 3.2a), one of the VSCs operates with the P-Q control scheme

where the active and reactive power errors are transformed into the reference d-q

current components, 𝑖𝑑∗ and 𝑖𝑞

∗ through the proportional-integral (PI) controllers.

Superscript asterisk denotes the reference values. The other VSC operates with the

𝑉𝑑𝑐 − 𝑄 control scheme maintaining a constant DC side voltage for a stable and

balanced active power flowing through the DC link. Dynamic limiters for 𝑖𝑑∗ and 𝑖𝑞

∗

are inserted to enable overcurrent limiting during network faults and disturbances. In

the inner current control loop (Figure 3.2b), the reference VSC d-q voltage, 𝑉𝑑𝑚 and

𝑉𝑞𝑚 are determined through the PI controllers considering the d-q current errors. The

voltage feed-forward and current feed-back compensations are used to get a good

dynamic response [86]. After transforming 𝑉𝑑𝑚 and 𝑉𝑞𝑚 into the VSC terminal

voltage by the inverse of Park’s transformation [87], the gate signals for the IGBTs are

obtained through the PWM. The PLL is important for the connection of VSCs to the

AC network in order to synchronize the phase of the output VSC voltage with the AC

network voltage. A PLL control topology based on the pq theory was used [88], as

shown in Figure 3.2c. By using the sum of the products of the feedback signals, 𝑓𝛼

and 𝑓𝛽 , and input 𝛼 − 𝛽 voltages (from Clark’s transformation [87] of the phase

voltages), the variation of the angular frequency ∆𝜔 is calculated as:

∆𝜔 = 𝑉𝛼 ∙ 𝑓𝛽 + 𝑉𝛽 ∙ 𝑓𝛼 (3-1)

The PLL output angle 𝜃 is then obtained using a PI-controller, a feedback of the base

angular frequency 𝜔𝑏 and an integrator.


40

+- PI

+- PI+- PIP

P*

Q

vdc

vdc*

iqmax

0

idmax

0

idmax

0

+- PI

Q*

Q

iqmax

0

id*

iq*

id*

iq*

Q*

Vdc -Q schemeP -Q scheme Power control loop

(a)

+PI+- +

+

-+

-+PI

wL

wLabc

dq

PWM

id

iq

vq

vd

id*

iq*

vqm

vdm vam vbm vcm

+

Gate signals

abc

dq

θ

θ

abc

dq

vd vq

Current control loop

vga

vgb vgc

iga

igb

igc

(b)

X

X

-sin

cos

θ

abc

ɑβ

PLL

w PI 1s

fɑ

fβ

∑ ∑

wb

w

vga vgb vgc

vɑ

vβ

(c)

Figure 3.2: Control block diagram of the SOP for power flow control mode: (a) outer

power control loop; (b) inner current control loop; (c) PLL controller

3.3.2 Supply Restoration Mode

When loads connected to one VSC of an SOP are isolated, the frequency and voltage of

this VSC are no longer dictated by the grid. Using the previous current control strategy

will cause voltage and/or frequency excursions and hence may lead to unacceptable

operating conditions [75].

In such a case, the interface VSC (connected to the isolated loads) changes to act as a


41

voltage source in order to provide a desired load voltage with stable frequency. The

other VSC still acts as a current source operating with the 𝑉𝑑𝑐 − 𝑄 control scheme.

Figure 3.3 shows the block diagram of the voltage and frequency control strategy for

the interface VSC. For voltage control, the VSC output voltage is controlled directly in

the d-q synchronous frame by holding 𝑉𝑞∗ to zero and controlling 𝑉𝑑

∗ as:

𝑉𝑑∗ = √2/3 ∙ 𝑉𝑟𝑚𝑠

∗ (3-2)

where 𝑉𝑟𝑚𝑠∗ is the desired nominal line to line rms voltage of the isolated loads.

√2/3 is included because Park’s transformation is based on the peak value of the

phase voltage. The output VSC voltage is regulated by closed loop control and

generated through the PWM scheme. For frequency control, a stable voltage frequency

is generated by using the PLL. The input phase voltages of the PLL are assigned by

transforming the reference d-q voltages, 𝑉𝑑∗ and 𝑉𝑞

∗ ( 𝑉𝑞∗ =0) through Clarke’s

transformation. Thus, ∆𝜔, as shown in Figure 3.2c, is calculated as:

∆𝜔 = 𝑉𝛼 ∙ 𝑓𝛽 + 𝑉𝛽 ∙ 𝑓𝛼 = 𝑉𝑑∗ ∙ cos 𝜃 ∙ (− sin 𝜃) + 𝑉𝑑

∗ ∙ sin 𝜃 ∙ cos 𝜃 = 0 (3-3)

Since ∆𝜔 remains zero, the phase angle 𝜃 with a fix frequency of 50 Hz can be

generated by integrating only the base angular frequency 𝜔𝑏 = 2𝜋 × 50 Hz.

+- abc

dq

PWM

Vrms*

PI

PIvam

vbm

vcm

+-

0

√2/3

abc

dq

abc

PLL

vla vlb vlc

θ

dq

vd*

vq*

vq

vd

vdm

vqm

Gate signals

θ

θ

θ

Figure 3.3: Control block diagram of the interface VSC for supply restoration control

mode


42

3.4 SOP Operation in MV Distribution Networks

The operating principle and performance of the back-to-back VSC based SOP under

different network operating conditions was analysed using a test system shown in

Figure 3.4. It consists of two MV feeders with constant impedance loads extracted from

an IEEE 33-bus distribution network [20] and a back-to-back VSC based SOP placed

between the feeders. This test system was modelled in EMTDC/PSCAD. The

parameters of the SOP device are shown in Table 3.1. An average model was used to

represent the SOP device, where losses, harmonics and fast switching transients of the

converters were neglected [89] [90].

SOP

VSC1 VSC2

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

018 19 20 21 22 23 24 25

HV/MV

Transformer

Figure 3.4: Two-feeder MV distribution network with a back-to-back VSC based SOP

Table 3.1: Parameters of the Back-to-back VSC Based SOP

Parameters Value

DC link voltage Vdc 35 kV

Filter inductor Lf 5 mH

DC link capacitor C 600 μF

Rated power 1 MVA

Rated VSC voltage (line to line) 12.66 kV/ 50 Hz

Rated VSC current 0.08 kA


43

3.4.1 Normal Conditions

The feasibility of the control strategy for the power flow control mode was verified.

Step changes on the active and reactive power reference signals were simulated. It was

assumed that VSC1 is operated in 𝑉𝑑𝑐-𝑄 and VSC2 is operated in 𝑃-𝑄 mode, as

shown in Figure 3.4. The reference step changes were simulated as follows: positive

and negative step changes on VSC1 reactive power 𝑄1∗ at 1s and 2s; on VSC2

reactive power 𝑄2∗ at 1.5s and 2.5s and active power 𝑃2

∗at 1.5s and 2s; the DC side

voltage 𝑉𝑑𝑐∗ remained a constant value.

Figure 3.5 shows the transient responses of the output power with respect to the step

changes in reference signals. Good tracking performance was observed with accurate

and rapid attainment of the steady state condition, i.e., attainted the steady state value

within a few milliseconds without either an under damped response or variation. These

results suggest the capability of the SOP to provide both instantaneous and longer

duration power flow regulation as well as dynamic Volt/VAR control under normal

network operating conditions. In addition, the DC link voltage remained a constant

value except transient variations occurring upon the step change in the active power, as

shown in Figure 3.5a. This is due to the step change of the active power which can

cause instantaneous imbalance of power exchange at the DC link.

Such undesired transient variations are usually damped by increasing the DC link

capacitance. However, it was found that when the step change in the active power

decreased, the variation in the DC voltage reduced. As shown in Figures 3.5a and 3.5c,

the DC voltage variation at t=2s caused by a step change of 0.5 MW was much smaller

than the variation at t=1.5s caused by a step change of 1 MW. This indicates that tuning

the active power controller (PI controller) for a slower transient (or step) response is

viable to alleviate undesirable DC voltage variation without enlarging the DC capacitor.


44

(a)

(b)

(c)

(d)

Figure 3.5: Transient response of the power flow control mode to step changes in

active and reactive power references: (a) DC side voltage; (b) reactive power response

of VSC1; (c) active power response of VSC2; (d) reactive power response of VSC2


45

3.4.2 During Fault Conditions

The behaviour of the SOP device and its controllers (power flow control mode) was

investigated under both balanced and unbalanced network faults. All simulations

presented in this section were conducted by assuming a fault occurs at Bus 13 (see

Figure 3.4). VSC1 is operated in 𝑉𝑑𝑐-𝑄 and VSC2 is operated in 𝑃-𝑄 mode.

Balanced fault on distribution network

A three-phase fault was simulated at t=1s with a fault impedance of 3 Ω. Such a high

impedance fault was chosen to avoid triggering the internal protection of the SOP

device that is typically designed to limit the VSC current to 2-3 times the full load

current [91]. It was assumed that the pre-fault direction of power flow is from the

un-faulted side to the faulted side feeder and the circuit breaker of the faulted side VSC

is tripped after 200ms.

Figure 3.6 presents the simulation results including the voltage, current and power flow

on both faulted and un-faulted side VSCs. Power was flowing from the un-faulted side

feeder to the faulted side feeder before the fault occurred. As seen in Figure 3.6a, the

faulted side VSC had a limited contribution to the fault current before tripping the

circuit breaker. The fault current became zero after tripping the circuit breaker.

However, as expected its output voltage dropped to 2kV (line to ground) and thus an

under-voltage protection scheme is sufficient to detect such a fault.

In contrast to the faulted side VSC, the un-faulted side VSC had a regulated output

current and voltage that were not affected by the fault, as shown in Figure 3.6b. This

suggests that the SOP device can effectively isolate a fault between the interconnected

feeders. A reduction in the output current is observed after the fault occurred. This is

due to the decrease of active power flow between the interconnected feeders.


46

Figure 3.6c shows the real and reactive power flow of both VSCs. Both active and

reactive power outputs on the faulted side VSC (dashed line) decreased during the fault,

and then became zero after tripping the circuit breaker. Nevertheless, the reactive

power output on the un-faulted side VSC (solid line) remained at the pre-fault value

due to the de-coupled controllability of active and reactive power.

(a)

(b)


47

(c)

Figure 3.6: SOP response for a three-phase fault at t=1s and blocked at t=1.2s: (a)

output voltage (left) and current (right) on the faulted side VSC; (b) output voltage (left)

and current (right) on the un-faulted side VSC; (c) power flow behaviour on both VSCs

When the pre-fault active power was flowing from the faulted side to the un-faulted

side feeder, a similar SOP response as shown in Figure 3.6 was observed

Unbalanced fault on distribution network

A single-phase to ground fault was simulated at t=1s and cut off by tripping the circuit

breaker after 200ms. The results are depicted in Figure 3.7. The faulted side VSC gave

a greater contribution to the fault current than in the balanced fault case. Therefore, the

network overcurrent protection is likely to react more quickly. Protection mechanisms

based on negative sequence current and voltage are sufficient to protect feeders with

SOP for this fault type. With respect to the un-faulted side VSC as shown in Figure

3.7b, it can be seen that both output current and voltage remained at the pre-fault values

despite the large fault current on the other side VSC.

Regarding the power flow during the unbalanced fault, the output active and reactive

power on the faulted side VSC remained at the pre-fault values but with notable ripple,


48

as depicted in Figure 3.7c. They became zero after the circuit breaker tripped. In

contrast, the reactive power supplied by the un-faulted side VSC was barely affected by

the fault, which remained at the pre-fault value. When the pre-fault active power flows

from the faulted side to the un-faulted side feeder, the SOP performance is similar to

those shown in Figure 3.7

(a)

(b)


49

(c)

Figure 3.7: SOP response for a single-phase to ground fault at t=1s and blocked at

t=1.2s: (a) output voltage (left) and current (right) on the faulted side VSC; (b) output

voltage (left) and current (right) on the un-faulted side VSC; (c) power flow behaviour

on both VSCs

Simulations have also been conducted when the faulted side VSC operates using the

𝑉𝑑𝑐-𝑄 scheme. Similar SOP performance was obtained to that shown in Figure 3.6 and

Figure 3.7. From the results obtained, it is concluded that the SOP device and its

controllers are effective in isolating the connected feeders during a fault, thus limiting

both the fault propagation on the network and the increase of short-circuit current. In

addition, despite the occurrence of a fault, the reactive power controllability on the

un-faulted side VSC remains.

3.4.3 Post-fault Supply Restoration Conditions

Under this condition, the SOP is required to provide a smooth transition between

different control modes: 1) from the power flow control mode to the supply restoration

control mode to resume power supply to the un-faulted out-of-service loads; and 2)

transfer back to the normal operation control mode when the isolated loads reconnect to

the original feeder. Figure 3.8 shows the transition system based on the control modes

proposed in Section 3.3. The d-q reference voltages input to the PWM modulation and


50

the input voltages of the PLL controller are changed by two switches ‘𝑆1’ and ‘𝑆2’. The

controllability of the transition system under the post-fault supply restoration

conditions was investigated. All simulations presented in this section follow the fault

occurrence described in Section 3.4.2. The feeder terminal units (FTUs) were assumed

to be installed in each branch and the total power of isolated loads were within the

capacity of the SOP device. Note that for a cold load pickup that exceeds the capacity

of the SOP device, more complicated restoration procedure should be considered.

However, this is beyond the scope of this work and will be considered as future

research.

+-

+-

PWM

abc

dq

+- Power

controller

vd,q*

vd,q

id,q

id,q* P

*/ vdc

*,Q

*

P / vdc , Q

vd,qm vam vbm vcm

Current

controller

Voltage

controller

Power flow controln mode

Supply restoration mode

SR

PFC

PLL

θ

vg(abc)

θ

dq

abc

vd*

0 SR

PFC

S1

S2

Figure 3.8: Control mode transition system

Transition from power flow control mode to supply restoration

mode

When a fault occurs at Bus 13 and is isolated by the network protection. The SOP is

changed to supply restoration mode to support the un-faulted out-of-service loads,

which is from Bus 14 to Bus 17. Once the supply restoration requirement is confirmed,

switches ‘𝑆1’ and ‘𝑆2’ simultaneously change from the power flow control mode, ‘PFC’

to the supply restoration mode, ‘SR’. It is assumed that the restoration requirement is

confirmed at t=2.2s, i.e., 1s after the fault was isolated. This response delay represents


51

the time used by the distribution automation to isolate a permanent fault and confirm

the requirement of supply restoration.

Figure 3.9 depicts the simulation results under a hard transition, i.e., directly connecting

‘𝑆1’ and ‘𝑆2’ to ‘SR’. It is observed that when the output voltage waveform of the

faulted side VSC was initiated from the desired reference value, a DC component was

produced in the output AC current, as illustrated in Figure 3.9b. This is due to the

inductive isolated loads.

(a)

(b)

Figure 3.9: Results of a hard transition to supply restoration mode at t=2.2s: (a) output

voltage waveform of the faulted side VSC; (b) output current waveform

In a highly inductive circuit, the current (sum of AC and DC currents) attempts to

remain the same upon an instantaneous change in the applied voltage [92]. Therefore,

when there is an instantaneous voltage change, the DC current component is produced

-14.0

-11.0

-8.0

-5.0

-2.0

1.0

4.0

7.0

10.0

13.0

2.1 2.15 2.2 2.25 2.3 2.35 2.4 2.45 2.5 2.55 2.6

Vo

ltag

e (k

V)

Time (s)

-0.06

-0.03

0.00

0.03

0.06

2.1 2.15 2.2 2.25 2.3 2.35 2.4 2.45 2.5 2.55 2.6

Cu

rren

t (k

A)

Time (s)


52

to compensate the change of AC current and thus lead to current DC offsets. As shown

in Figure 3.10a, the current DC offsets became more severe when an inductance was

connected in parallel with the isolated load (at Bus 17). Moreover, when a capacitance

was connected in parallel with the isolated load (at Bus 17), a transient impulse was

observed in the output current waveform, as shown in Figure 3.10b. The capacitive

loads such as power factor correction and voltage support capacitors connected to the

distribution network are normally not automatically disconnected from the network

during fault isolation. Both of these current DC offset and transient impulse cannot be

ignored due to the probability of triggering the inverter internal protection as well as

the feeder protective relays.

(a)

(b)

Figure 3.10: Output current waveforms of the faulted side VSC for a hard transition to

supply restoration mode with: (a) an inductance connecting to Bus 17; (b) a

capacitance connecting to Bus 17

-0.04

-0.03

-0.01

0.01

0.02

0.04

2.1 2.15 2.2 2.25 2.3 2.35 2.4 2.45 2.5 2.55 2.6

Cu

rren

t (k

A)

Time (s)

-0.14

-0.10

-0.06

-0.02

0.02

0.06

0.10

0.14

2.1 2.15 2.2 2.25 2.3 2.35 2.4 2.45 2.5 2.55 2.6

Cu

rren

t (k

A)

Time (s)


53

A soft cold load pickup process to solve the current DC offset or inrush problems was

proposed:

1) To avoid a sudden jump of the terminal voltage, at the instant of connecting

switches ‘𝑆1’ and ‘𝑆2’ to ‘SR’, the initial values of both PLL angle and output

voltage of the voltage controller are set to the values of the PLL angle and

voltage magnitude just before switching to ‘SR’.

2) The voltage controller (PI controller) is set for a slower transient response, i.e.,

a larger time constant to achieve a soft increase in the output voltage.

Figure 3.11 shows the output voltage and current waveforms after using the soft cold

load pickup process. It can be seen that the output voltage was increased smoothly

without a sudden jump to the reference value. The DC offset in the output current was

successfully removed.

(a)

(b)

Figure 3.11: Results of a smooth transition to supply restoration mode at t=2.2s: (a)

output voltage waveform of the faulted side VSC; (b) output current waveform

-14.0

-11.0

-8.0

-5.0

-2.0

1.0

4.0

7.0

10.0

13.0

2.1 2.15 2.2 2.25 2.3 2.35 2.4 2.45 2.5 2.55 2.6

Vo

ltag

e (k

V)

Time (s)

-0.06

-0.03

0.00

0.03

0.06

2.1 2.15 2.2 2.25 2.3 2.35 2.4 2.45 2.5 2.55 2.6

Cu

rren

t (k

A)

Time (s)


54

Transition from supply restoration mode to power flow control

mode

After the fault clearance, the isolated loads supplied by the SOP are reconnected to the

original feeder. The SOP has to be changed from the supply restoration mode to the

power flow control mode by switching ‘𝑆1’ and ‘𝑆2’. The reconnection was assumed to

occur at t=3s. Figure 3.12 depicts the simulation results under a hard transition, i.e.,

directly connecting ‘𝑆1’ and ‘𝑆2’ to ‘PFC’. There was a current spike and voltage sag

upon the transition, which may result in unexpected operation of the protection relays.

(a)

(b)

Figure 3.12: Results of a hard transition to power flow control mode at t=3s: (a) output


The reconnection is carried out after both voltage phase angle and magnitude are

synchronized with the original feeder in order to avoid excessive transients. The

detailed procedure for a seamless transition from the supply restoration mode to the

power flow control mode is illustrated as follows:

-13.0

-10.0

-7.0

-4.0

-1.0

2.0

5.0

8.0

11.0

2.8 2.85 2.9 2.95 3 3.05 3.1 3.15 3.2

Vo

ltag

e (k

V)

Time (s)

-0.20

-0.15

-0.10

-0.05

0.00

0.05

0.10

0.15

0.20

2.8 2.85 2.9 2.95 3 3.05 3.1 3.15 3.2

Cu

rren

t (k

A)

Time (s)


55

1) Confirm that reconnection is required after the network fault has been removed.

2) The phase of the load voltage, 𝑉𝑓_𝑟 at the reconnection point, i.e., at Bus 14,

is synchronized with that of the original feeder voltage at Bus 13, 𝑉𝑙_𝑟:

The phase difference between the original feeder and the reconnection point

voltage is obtained by

∆𝜃 = 𝑉𝑓_𝑟 − 𝑉𝑙_𝑟 (3-4)

Two sets of voltage values are used to obtain the information ∆𝜃

𝑎 = 𝑉𝑓_𝑟𝑎 ∙ 𝑉𝑙_𝑟𝑎 + 𝑉𝑓_𝑟𝑏 ∙ 𝑉𝑙_𝑟𝑏 + 𝑉𝑓_𝑟𝑐 ∙ 𝑉𝑙_𝑟𝑐 =3

2cos ∆𝜃 (3-5)

𝑏 = 𝑉𝑙_𝑟𝑎 ∙ 𝑉𝑓_𝑟𝑏 + 𝑉𝑙_𝑟𝑏 ∙ 𝑉𝑓_𝑟𝑐 + 𝑉𝑙_𝑟𝑐 ∙ 𝑉𝑓_𝑟𝑎

=3

4[− cos ∆𝜃 + √3 sin ∆𝜃] (3-6)

Combining (3-5) and (3-6), sin ∆𝜃 can be found as

sin ∆𝜃 =√3

3∙ (

4

3𝑏 +

2

3𝑎) (3-7)

Figure 3.13 shows an overview of the phase synchronisation process. ‘S2’ is

switched to ‘PFC’ after the phase synchronisation is completed.

Eq. (3-5)

Eq. (3-6)

vf_ra vf_rb vf_rc

a

bEq. (3-7)

sin( θ)PI

++

dq

abc

vd*

0 PLL

θ

vl_ra vl_rb vl_rc

θold

θnew

θold

Figure 3.13: Synchronisation controller

3) The d−q feeder voltages are assigned as the d−q voltage references of the

voltage controller in order to adjust the magnitude of the load voltage to that of


56

the grid voltage.

4) When both phase and magnitude of the output voltage match the original feeder

voltage, ‘S1’ is switched to ‘PFC’, and the isolated loads are reconnected to the

original feeder. To prevent large transient flows of real and reactive power, the

integers in both power and current controllers are reset and the reference values

of real and reactive power are set as the value before transfer to the power flow

control mode.

Figure 3.14 shows the synchronisation of the voltages when the synchronisation

algorithm started to work in the supply restoration mode. As can be seen, the

synchronisation process successfully forced the voltage at the end of the isolated area to

track the voltage at the original feeder.

(a)

(b)

Figure 3.14: Voltage synchronisations for reconnection: (a) phase synchronisation; (b)

magnitude synchronisation

-12.0

-8.0

-4.0

0.0

4.0

8.0

12.0

2.48 2.5 2.52 2.54 2.56 2.58 2.6

Vo

ltag

e (k

V)

Time (s)

Vf_ra

Vl_ra

0.0

4.0

8.0

12.0

2.6 2.62 2.64 2.66 2.68 2.7 2.72

Vo

ltag

e (k

V)

Time (s)

Vf_r

Vl_r


57

Figure 3.15 depicts the output voltage and current waveforms of the faulted side VSC

after using the synchronisation process. The current spike and voltage sag were

successfully reduced.

(a)

(b)

Figure 3.15: Results of a smooth transition to power flow control mode: (a) output


3.5 Summary

Two control modes were developed for the operation of a back-to-back VSC based SOP.

For the power flow control mode, a dual closed-loop current-controlled strategy was

used to regulate both active and reactive power on the connected feeders. For the

supply restoration mode, a voltage and frequency control strategy was proposed to

-13.0

-10.0

-7.0

-4.0

-1.0

2.0

5.0

8.0

11.0

2.8 2.85 2.9 2.95 3 3.05 3.1 3.15 3.2

Vo

ltag

e (k

V)

Time (s)

-0.06

-0.04

-0.02

0.00

0.02

0.04

0.06

2.8 2.85 2.9 2.95 3 3.05 3.1 3.15 3.2

Cu

rren

t (k

A)

Time (s)


58

provide fast power supply restoration when the loads connected to one side of SOP are

isolated due to network faults. The operational principle of the back-to-back VSC based

SOP was analysed under both normal and abnormal network operating conditions. Case

studies based on a two-feeder MV distribution network evaluated the performance of

the SOP device with the two control modes under normal, faults and post-fault supply

restoration conditions.

Under normal conditions, the proposed SOP controller provides both instantaneous and

longer duration power flow regulation between the interconnected feeders and dynamic

Volt/VAR support on both terminals.

During fault conditions, the SOP controller (power flow control mode) is able to isolate

both balanced and unbalanced network faults between interconnected feeders. The SOP

is therefore capable of limiting both the fault propagation on the network and the

increase of the short-circuit current.

Under post-fault supply restoration conditions, the SOP device was demonstrated to be

effective in providing fast supply restoration through a smooth transition between the

power flow control and the supply restoration modes. For the transition from the power

flow control mode to the supply restoration mode, a soft cold load pickup process was

preferred over the hard transition. The soft cold load pickup process removed the DC

offset and inrush problems of the output current which may trigger unexpected

protection operations. For the transition from the supply restoration mode to the power

flow control mode, the undesirable current impulse and voltage drop were reduced after

the use of the voltage synchronisation procedure and the reset of current and power

controllers.

Chapter 4 Benefits Analysis of SOPs for MV Distribution Network Operation

59

Benefits Analysis of SOPs for

MV Distribution Network

Operation

HIS chapter investigates the operational benefits of SOPs for MV

distribution network operation including power loss minimization,

feeder load balancing and voltage profile improvement.

Chapter 4

T


60

4.1 Introduction

In this chapter, a steady state analysis framework was developed to quantify the

operational benefits of a MV distribution network with SOPs.

A generic power injection model of an SOP for steady state analysis was developed,

which takes into account both physical limitations and internal power losses of the SOP

device, using back-to-back VSCs. Based on the SOP model, an improved Powell’s

direct set method was developed to obtain the optimal SOP operation. Distribution

network reconfiguration algorithms, with and without SOPs, were also developed and

used to identify the benefits of using SOPs.

The performance of using SOPs, traditional network reconfiguration and the

combination of both was compared and analysed based on a 33-bus distribution

network. The impact of DG connections and the impact of SOP device losses were also

evaluated through quantitative and sensitive analysis.

4.2 Modelling of Soft Open Points

Figure 4.1a shows a typical location of an SOP which allows power electronic devices

to control active power flow between connected feeders and to supply or absorb

reactive power at its interface terminals under normal network operating conditions. A

generic power injection model of SOP was developed. This model considers SOP

terminal power injections and hence enables the incorporation of SOPs into existing

power flow analysis algorithms without considering the detailed controller design.


61

(a)

(b)

Figure 4.1: (a) Simple distribution network with an SOP; (b) Power injection model of

SOP for distribution network power flow control

Figure 4.1b shows the representation of an SOP model with real and reactive power,

injecting into feeders I and J through both terminals. Taking these power injections as

decision variables, the power flow in feeder I is calculated by the following set of

recursive equations [93]:

𝑃𝑖 = 𝑃𝑖−1 − 𝑃𝐿𝑜𝑠𝑠 (𝑖−1, 𝑖) − 𝑃𝐿,𝑖 = 𝑃𝑖−1 −𝑟𝑖−1

|𝑉𝑖−1|2∙ (𝑃𝑖−1

2 + 𝑄𝑖−12) − 𝑃𝐿,i (4-1)

𝑄𝑖 = 𝑄𝑖−1 − 𝑄𝐿𝑜𝑠𝑠 (𝑖−1, 𝑖) − 𝑄𝐿,𝑖 = 𝑄𝑖−1 −𝑥𝑖−1

|𝑉𝑖−1|2∙ (𝑃𝑖−1

2 + 𝑄𝑖−12) − 𝑄𝐿,i (4-2)

|𝑉𝑖|2 = |𝑉𝑖−1|2 − 2 ∙ (𝑟𝑖−1𝑃𝑖−1 + 𝑥𝑖𝑄𝑖) +

(𝑟𝑖−12+𝑥𝑖−1

2)

|𝑉𝑖−1|2∙ (𝑃𝑖−1

2 + 𝑄𝑖−12) (4-3)


62

with boundary conditions:

|𝑃𝑛 + 𝑃𝑆_𝑖𝑛𝑗𝐼 | = 𝑃𝐿𝑜𝑠𝑠 (𝑛, 𝑠𝐼) (4-4)

|𝑄𝑛 + 𝑄𝑆_𝑖𝑛𝑗𝐼 | = 𝑄𝐿𝑜𝑠𝑠 (𝑛, 𝑠𝐼) (4-5)

where 𝑃 represents active power, 𝑄 reactive power, 𝑉 nodal voltage, 𝑟 resistance

and 𝑥 reactance. Subscript 𝐿𝑜𝑠𝑠 denotes line losses, and 𝐿 denotes load. The

variables and parameters are shown in Figure 4.1b. Similar recursive power flow

equations with boundary conditions are applied to feeder J.

To consider the internal losses of the SOP equipment, the following equality constraint

of power balance is used:

𝑃𝑆_𝑖𝑛𝑗𝐼 + 𝑃𝑆_𝑖𝑛𝑗

𝐽 + 𝑃𝑆𝑂𝑃,𝐿𝑜𝑠𝑠 = 0 (4-6)

where 𝑃𝑆𝑂𝑃,𝐿𝑜𝑠𝑠 denotes the internal power losses of the whole SOP device.

Various types of power electronic devices can be implemented as an SOP, such as

UPFC, back-to-back VSCs and multi-terminal VSCs. In this work, back-to-back VSCs

were used. Constraints of such device due to the physical limitations and the internal

power losses are formulated as follows.

4.2.1 Physical Limitations of Back-to-Back Converters

As described in Section 2.3, back-to-back VSCs consist of two VSCs connected via a

DC link. Both VSCs are able to provide fast and independent control of active and

reactive power in all four quadrants of the P-Q plane. The terminal power injections of

back-to-back VSCs operated as a SOP are controlled directly by each VSC, the

operational limits of VSC capacity and terminal voltage are considered as:


63

√ 𝑃𝑆_𝑖𝑛𝑗𝐼 2

+ 𝑄𝑆_𝑖𝑛𝑗𝐼 2

≤ 𝑆𝑉𝑆𝐶,𝑟𝑎𝑡𝑒𝐼 (4-7)

𝑉𝑠𝐼 ≤ 𝑉𝑉𝑆𝐶,𝑟𝑎𝑡𝑒𝐼 (4-8)

where 𝑆𝑉𝑆𝐶,𝑟𝑎𝑡𝑒𝐼 is the power rating of the VSC connected to feeder I; 𝑉𝑠𝐼

and 𝑉𝑉𝑆𝐶,𝑟𝑎𝑡𝑒𝐼 denote the actual and maximum AC terminal voltage. Similar constraints

are applied to the other side VSC connected to feeder J.

4.2.2 Internal Power Losses of Back-to-Back Converter

Internal power losses of the back-to-back device are made up of the losses of its

individual components, including the semiconductors (conduction and switching

losses), passive components (DC link capacitor, filter, AC line choke), transformers and

the cooling system.

As suggested in [94], these losses can be categorized into three components: no load

losses, linear and quadratic losses depending on the converter current, which is a

function of the active and reactive power exchanged with the AC network:

𝑃𝑉𝑆𝐶,𝐿𝑜𝑠𝑠𝐼 = 𝑎𝐼 ∙ 𝐼𝑉𝑆𝐶

𝐼 2+ 𝑏𝐼 ∙ 𝐼𝑉𝑆𝐶

𝐼 + 𝑐𝐼 (4-9)

𝐼𝑉𝑆𝐶𝐼 =

√ 𝑃𝑆_𝑖𝑛𝑗𝐼 2

+ 𝑄𝑆_𝑖𝑛𝑗𝐼 2

|𝑉𝑠𝐼| (4-10)

where 𝑃𝑉𝑆𝐶,𝐿𝑜𝑠𝑠𝐼 and 𝐼𝑉𝑆𝐶

𝐼 represent the power losses and the AC current of the VSC

connecting to feeder I. Similar equations are applied to the other side VSC. Thus all

equations shown in the following section only illustrate the VSC connected to feeder I.

As shown in (4-9), the total losses are determined by the coefficients, 𝑎𝐼, 𝑏𝐼and 𝑐𝐼.

These coefficients are difficult to obtain due to the limited information from the open

literature or manufacturers. Usually only the internal power losses (or efficiency) under


64

nominal conditions (rated power) and the no load loss, 𝑐𝐼are given.

A linear function is used to approximate the quadratic one in (4-9):

𝑃𝑉𝑆𝐶,𝐿𝑜𝑠𝑠𝐼 = 𝑘𝐼 ∙ 𝐼𝑉𝑆𝐶

𝐼 + 𝑐𝐼 (4-11)

𝑘𝐼 = (𝑃𝑉𝑆𝐶,𝐿𝑜𝑠𝑠,𝑟𝑎𝑡𝑒𝐼 − 𝑐𝐼)/𝐼𝑉𝑆𝐶,𝑟𝑎𝑡𝑒

𝐼 (4-12)

where 𝑃𝑉𝑆𝐶,𝐿𝑜𝑠𝑠,𝑟𝑎𝑡𝑒𝐼 and 𝐼𝑉𝑆𝐶,𝑟𝑎𝑡𝑒

𝐼 are power losses and AC current of the VSC

connected to feeder I under nominal condition.

Comparing the internal power losses derived from (4-9) and (4-11), as shown in Figure

4.2, it is noted that when 𝐼𝑉𝑆𝐶𝐼 ≤ 𝐼𝑉𝑆𝐶,𝑟𝑎𝑡𝑒

𝐼 , the device power losses using the

approximate model are higher than those using the quadratic one, i.e., more

conservative conclusions are drawn. Therefore, the approximate model was used in this

work.

Figure 4.2: Comparison between the quadratic and approximate loss estimation

function


65

4.3 Optimal Operation of the Soft Open Points

The benefits of SOPs for both system loss reduction and feeder load balancing were

investigated. To quantify these benefits, the amounts of real and reactive power

injections of the SOPs were determined through solving a combinational nonlinear

constrained optimization problem.

4.3.1 Problem Formulation

Two objective functions are used separately to quantify the optimality of the SOP

operation.

Active power loss minimization

The active power losses consist of two components: the feeder losses and the internal

power losses of the SOP devices. The active power loss minimization problem is

formulated as:

Minimize 𝑃𝑇,𝐿𝑜𝑠𝑠 = ∑ 𝑟𝑘 ∙𝑃𝑘

2+𝑄𝑘2

|𝑉𝑘|2

𝑛𝑙𝑘=1 + 𝑃𝑆𝑂𝑃,𝐿𝑜𝑠𝑠 (4-13)

where 𝑟𝑘 , 𝑃𝑘 , 𝑄𝑘 , 𝑉𝑘 are resistance, active power, reactive power, and voltage of

branch k; and 𝑛𝑙 is the total number of branches in the network.

Feeder load balancing

A branch loading index 𝐿𝐼𝑘 is used to measure the loading level of each branch in the

network, which is expressed as:

𝐿𝐼𝑘 = (𝐼𝑘

𝐼𝑘,𝑟𝑎𝑡𝑒)

2

∀ 𝑘 ∈ 𝑛𝑙 (4-14)

where 𝐼𝑘 and 𝐼𝑘,𝑟𝑎𝑡𝑒 are the actual and the rated branch current of branch k.

Feeder load balancing is achieved by minimizing the load balance index 𝐿𝐵𝐼, which is

defined as the sum of branch load balancing indices, as shown in (4-15):


66

𝑀𝑖𝑛𝑖𝑚𝑖𝑧𝑒 𝐿𝐵𝐼 = ∑ 𝐿𝐼𝑘𝑛𝑙𝑘=1 (4-15)

Constraints

Constraints of the distribution network and the SOP devices are considered. The device

constraints of each SOP are shown in (4-6)-(4-8). The network constraints are

expressed as follows:

𝑔(𝑥, 𝑠) = 0 (4-16)

|𝑉𝑖,𝑚𝑖𝑛| ≤ |𝑉𝑖| ≤ |𝑉𝑖,𝑚𝑎𝑥| ∀ 𝑖 ∈ 𝑛𝑏 (4-17)

|𝑆𝑘| ≤ |𝑆𝑘,𝑚𝑎𝑥| ∀ 𝑘 ∈ 𝑛𝑙 (4-18)

where 𝑔(𝑥, 𝑠), 𝑉𝑖,𝑚𝑖𝑛, 𝑉𝑖,𝑚𝑎𝑥 and 𝑆𝑘,𝑚𝑎𝑥 are the power flow equations, the minimum

and the maximum voltage of bus i and the maximum capacity allowed in branch k. 𝑛𝑏

are the total number of buses of the network.

Penalty function

Constrains of mentioned above are included into the objective function by using the

penalty function method. In this way, unconstrained optimization methods are used

directly by solving the transformed objective function (with penalty terms):

𝑀𝑖𝑛𝑖𝑚𝑖𝑧𝑒 𝐹𝑜𝑏𝑗(𝑺) = {𝑃𝑇,𝐿𝑜𝑠𝑠 𝑜𝑟 𝐿𝐵𝐼} + 𝑘1 ∙ 𝑓𝑆𝑂𝑃 + 𝑘2 ∙ 𝑓𝑉 + 𝑘3 ∙ 𝑓𝑆 (4-19)

where 𝑘1, 𝑘2, 𝑘3 are the penalty constants; 𝑓𝑆𝑂𝑃 , 𝑓𝑉 , and 𝑓𝑆 are the penalty functions

for violations of the SOP device constraints, the network voltage and capacity

constraints; S denotes the decision variable vector.

Based on the power injection model of SOP in Section 4.2, the real and reactive power

injections at both terminals are specified as the decision variables: with one SOP

installation, 𝑺 = [𝑃𝑆𝑖𝑛𝑗

𝐼 , 𝑄𝑆𝑖𝑛𝑗

𝐼, 𝑄𝑆𝑖𝑛𝑗

𝐽 ]T. 𝑃𝑆_𝑖𝑛𝑗𝐽 is not included in 𝑺 since it is determined if


67

𝑺 is specified according to the equality constraint equation of (4-6). Therefore, the

coordinates of n SOPs are combined together sequentially to obtain a point of 𝑺 in a

3n-dimentional space, thus

𝑺 = [𝑃𝑆_𝑖𝑛𝑗, 1𝐼 , 𝑄𝑆_𝑖𝑛𝑗, 1

𝐼, 𝑄𝑆_𝑖𝑛𝑗, 1

𝐽, 𝑃𝑆_𝑖𝑛𝑗, 2

𝐼 , 𝑄𝑆_𝑖𝑛𝑗, 2𝐼

, 𝑄𝑆_𝑖𝑛𝑗, 2𝐽

,

… , 𝑃𝑆𝑖𝑛𝑗, 𝑛𝐼 , 𝑄𝑆𝑖𝑛𝑗, 𝑛

𝐼 , 𝑄𝑆𝑖𝑛𝑗, 𝑛𝐽 ]T (4-20)

4.3.2 Method of Determining Optimal SOP Operation

Based on the Powell’s Direct Set (PDS) method presented in [95], an improved PDS

method to optimize the SOP operation was developed. The performance improvement

is achieved by obtaining a good initial approximated SOP operation.

Powell’s Direct Set method

Most mathematical optimization methods require expressions of the derivatives to

define the direction of movement, i.e. search the direction approaching to the optimum

(from a starting point).

The Powell’s Direct Set method is a direct search method proposed by Powell et al.

[95]. It defines the search directions in a direct manner, i.e., solely depending on the

objective function itself. Therefore, this method is easy to implement and is not limited

by the existence of derivatives of the objective function. It has been successfully

applied to solve problems for which it is difficult or impossible to calculate the

derivatives [96-98].

A comprehensive review of PDS method as well as the mathematical proof of its

convergence were given in [95]. Three key properties of this method are highlighted

below:


68

i. For an N-dimensional problem, minimization is achieved by an iterative procedure

that searches down N linear independent directions within each iteration, i.e.

starting from the best known approximation to the optimum.

ii. Fast convergence to the optimum is achievable by only searching down N mutually

conjugate directions. The optimal solution of a quadratic function has been proved

achievable by searching along those mutually conjugate directions once only.

Hence only N iterations are required to solve the N-dimensional quadratic problem

[95]. The efficacy has also been demonstrated for any other function forms, as

presented in [96, 97], but may require more iterations.

iii. The mutually conjugate directions are generated after each iteration, as illustrated

in the following part.

PDS method for optimal SOP operation

The process of determining optimal SOP operation using the PDS method is shown in

Figure 4. 3.

(1) Initialization

(2) Generate a new mutually conjugate direction ƺConju.

(4) Find the minimum of Fobj along the conjugate direction ƺConju.

(5)Stopping rule satisfied?

(6) Output the results:

Fobj and SOP operation S;

(7) End

No

Yes

(3) Update the search direction set {ƺ} ?

Yes

No

Figure 4.3: Flow chart of the proposed PDS method


69

STEP 1): Initialization. Based on (4-20), the initial approximate SOP operation 𝑺𝟎(1)

,

the initial search direction set with 3n linear independent search directions, {𝝃}(𝟏) =

{ 𝝃𝟏(1)

, … , 𝝃𝒊(1)

, … , 𝝃𝟑𝒏(1)

} and the convergence criterion 𝜀 are specified. The 3n

directions in {𝝃}(𝟏) are initially chosen to be the co-ordinate directions (linear

independent). This means only one decision variable in (4-20) will be changed when

searching along one direction.

STEP 2): Generate a new mutually conjugate direction within one iteration. There are

two sub steps:

i. Starting from 𝑺𝟎(𝑘)

, k indicates the iteration number (one iteration includes

searching along 3n directions, k=1 initially), sequentially find 𝑺𝒊(𝒌)

which gives

the minimum of the objective function (4-19) along each search direction, 𝝃𝒊(𝑘)

in {𝝃}(𝒌), which is expressed as:

Min 𝐹𝑜𝑏𝑗(𝑺𝒊(𝑘)

) = 𝐹𝑜𝑏𝑗(𝑺𝒊−𝟏(𝑘)

+ 𝜆𝑖 ∙ 𝝃𝒊(𝑘)

), i=1, 2,…, 3n (4-21)

In this work, a one-dimensional search method, the golden section search [95],

was adopted to calculate the optimal step size, 𝜆𝑖. The load flow analysis method

introduced in [99] combined with the SOP injection model formulated in (4-1) -

(4-6) was implemented as a subroutine to calculate the objective function

𝐹𝑜𝑏𝑗(𝑺𝒊(𝑘)

) in (4-21).

ii. After searching down the 3n directions, a conjugate direction is generated by

(4-22)

𝝃𝒄𝒐𝒏𝒋𝒖.(𝑘)

= 𝑺𝟑𝒏(𝑘)

− 𝑺𝟎(𝑘)

(4-22)


70

STEP 3): Update the search direction set for the next iteration {𝝃}(𝒌+𝟏). Two scenarios

are considered:

i. If n=1, discard the first direction 𝝃𝟏(𝑘)

in { 𝝃𝒊(𝑘)

} by adding 𝝃𝒄𝒐𝒏𝒋𝒖.(𝑘)

to the end,

as shown in (4-23):

{𝝃}(𝒌+𝟏) = { 𝝃𝟐(𝑘)

, 𝝃𝟑(𝑘)

, … , 𝝃𝟑𝒏(𝑘)

, 𝝃𝒄𝒐𝒏𝒋𝒖.(𝑘)

} (4-23)

ii. If n>1, a ‘smarter’ updating procedure is required to ensure a reasonable rate of

convergence, i.e., replace the direction in {𝝃}(𝒌) , which shows the worst

performance. The general procedure is presented as follows [95]:

Find 𝝃𝒎(𝑘)

that gives maximum reduction among the previous

one-dimensional searching processes in Step 2, as shown in (4-24):

∆𝑚𝑎𝑥= 𝑀𝑎𝑥𝟏≤𝒎≤𝟑𝒏

{ 𝐹𝑜𝑏𝑗(𝑺𝒎−𝟏(𝑘)

) − 𝐹𝑜𝑏𝑗(𝑺𝒎(𝑘)

)} (4-24)

Replace 𝝃𝒎(𝑘)

by 𝝃𝒄𝒐𝒏𝒋𝒖.(𝑘)

to give more efficient convergence if the

following two criterions are satisfied:

(𝑤1 − 2𝑤2 + 𝑤3)(𝑤1 − 𝑤2 − ∆𝑚𝑎𝑥)2 < 0.5 ∙ ∆𝑚𝑎𝑥(𝑤1 − 𝑤3)2

𝑤3 < 𝑤1} (4-25)

where 𝑤1 = 𝐹𝑜𝑏𝑗(𝑺𝟎(𝑘)

), 𝑤2 = 𝐹𝑜𝑏𝑗 (𝑺𝟑𝒏(𝑘)

),𝑤3 = 𝐹𝑜𝑏𝑗 (2 ∙ 𝑺𝟑𝒏(𝑘)

− 𝑺𝟎(𝑘)

). Thus,

{𝝃}(𝒌+𝟏) = { 𝝃𝟏(𝑘)

, … , 𝝃𝒎−𝟏(𝑘)

, 𝝃𝒄𝒐𝒏𝒋𝒖.(𝑘)

, 𝝃𝒎+𝟏(𝑘)

, … , 𝝃𝟑𝒏(𝑘)

} (4-26)

Otherwise{𝝃}(𝒌)is not updated in this iteration, let 𝑺𝟎(𝑘+1)

= 𝑺𝟑𝒏(𝑘)

and jump

to Step 5.


71

STEP 4): Search down along the conjugate direction. Calculate 𝜆 as (4-21) so

that 𝐹𝑜𝑏𝑗( 𝑺𝟎(𝑘+1)

) reaches its minimum along the conjugate direction generated in

(4-22). Let 𝑺𝟎(𝑘+1)

= 𝑺𝟑𝒏(𝑘)

+ 𝜆 ∙ 𝝃𝒄𝒐𝒏𝒋𝒖.(𝑘)

.

STEP 5): Convergence check. If ‖ 𝑺𝟎(𝑘+1)

− 𝑺𝟎(𝑘)

‖ > 𝜀, let k=k+1 and then go to Step 2,

else stop.

An example shown in Figure 4.4 illustrates the searching of the optimization process.

As the red solid line shows, the optimization starts from an initial approximation 𝑺𝟎(1)

and then reaches the minimum 𝑺𝟏(1)

, 𝑺𝟐(1)

, 𝑺𝟑(1)

along each of the coordinate directions

in {𝝃}(𝟏). The first conjugate direction 𝝃𝒄𝒐𝒏𝒋𝒖.(1)

is hence obtained, as shown by the blue

dotted arrow. After searching along the conjugate direction, a new approximation 𝑺𝟎(2)

is obtained for the second iteration.

Figure 4.4: Example of one SOP optimization by PDS method


72

Determine the initial approximated SOP operation

The PDS method is able to start from any initial point 𝑺𝟎(1)

. A good approximation of

SOP operation is obtained by running the PDS method once using the objective

function of (4-21) with simplified power flow equations given by [99]:

i. All nodal voltage magnitudes are assumed to be 1 p.u., equations (4-13) and

(4-15) are reduced to

𝑃𝐿𝑜𝑠𝑠 ≈ ∑ 𝑟𝑘 ∙ (𝑃𝑘2 + 𝑄𝑘

2) + 𝑃𝑆𝑂𝑃,𝐿𝑜𝑠𝑠𝑛𝑙𝑘=1 (4-27)

𝐿𝐼𝐵 ≈ ∑𝑃𝑘

2+𝑄𝑘2

𝐼𝑘,𝑟𝑎𝑡𝑒2

𝑛𝑙𝑘=1 (4-28)

ii. Ignoring all the quadratic terms (line losses) from the power flow

equations in (4-1)-(4-3). The branch power 𝑃𝑘 and 𝑄𝑘 are obtained by

summing up the downstream power loads:

𝑃𝑘 ≈ ∑ 𝑃𝐿,𝑖𝑛𝑖=𝑘+1 (4-29)

𝑄𝑘 ≈ ∑ 𝑄𝐿,𝑖𝑛𝑖=𝑘+1 (4-30)

The constraints of nodal voltage and branch capacity are considered by:

|𝑉𝑖|2 ≈ |𝑉0|2 − 2 ∑ (𝑟𝑘𝑃𝑘 + 𝑥𝑘𝑄𝑘) ≤ 𝑉𝑚𝑎𝑥

2𝑖𝑘=1 (4-31)

𝑆𝑘 ≈ (𝑃𝑘2 + 𝑄𝑘

2) ≤ 𝑆𝑘,𝑚𝑎𝑥2 (4-32)

Using the simplified equations, as shown in (4-27)-(4-32), the initial approximated SOP

operation is able to be obtained directly without using the accurate load flow

calculations.


73

4.4 Network Reconfiguration Considering SOPs

Reconfiguring distribution networks is used to achieve better network operation

including power loss minimization, feeder load balancing and supply restoration. To

investigate the performance of distribution network reconfiguration when SOPs are

installed to replace some of the normally-open points, the proposed PDS method for

optimal SOP operation was combined with the network reconfiguration method

introduced in [34].

The procedure of the combined algorithm is given in Figure 4.5.

Start

Determine the sequence of load busbars to search supply paths

Find the optimal network configuration using a shortest path search algorithm

Calculate the combined results as fitness values

Stopping criteria is satisfied?

Output optimization results

End

GA algorithm

with

Selection,

crossover and

mutation

Yes

No

Determine optimal SNOP operation using the PDS method

Figure 4.5: Flowchart of proposed network reconfiguration considering SOPs


74

According to [34], a shortest-path algorithm is used to find the optimal electricity

supply path for each load busbar. A genetic algorithm (GA) with the selection,

crossover and mutation operators is used to optimize the sequence of load busbars

searching for the supply paths because the sequence affects the obtained network

configurations. The improved PDS method for optimal SOP operation is integrated

after the shortest path algorithm. The fitness functions are formulated in Section 4.3.1.

4.5 Case Study

A 33-bus distribution network, as shown in Figure 4.6, was used for case study [99].

This network has 32 normally-closed switches, 5 normally-open switches and the

nominal voltage is 12.66 kV. The total real and reactive power loads are 3,715 kW and

2,300 kVar. Four normally-open switches, i.e., the switches between buses 25 and 29,

33 and 18, 8 and 21, 12 and 22, were chosen as candidate places for SOP installation.

The size of each SOP unit was assumed to be 3 MVA. Such a large size is chosen to

ensure enough capacity for SOP operations under all studied network conditions.

Figure 4.6: 33-bus distribution network [99]

Four cases were defined and used for quantifying the benefits of SOPs for improvement

of network performance:


75

Case I: Improve network performance using SOPs;

Case II: Improve network performance considering both SOPs and network

reconfiguration;

Case III: Impacts of DG connections;

Case IV: Impact of power losses of SOP devices.

4.5.1 Improve Network Performances Using SOPs

The impact of different number of SOPs installed in the network on both power loss

minimization and feeder load balancing was investigated. The device power losses

were ignored for this case.

Power loss minimization

Figure 4.7a shows the results of minimized power losses with 1 to 4 SOPs installed in

the network. There was a significant reduction (by 42%) of the system power loss with

one SOP installation. With more SOPs installed in the system, the total power losses

were further reduced (by 57%).The voltage profiles of the system were also improved

with the SOP installations, as shown in Figure 4.7b.

(a)

0

60

120

180

240

0 1 2 3 4

Tota

l Po

wer

Lo

sses

(kW

)

Number of SOP

Effect of Number of SOP Installation on Power Loss

No. SOP Locations (Branches)1 25-292 25-29 18-333 25-29 18-33 12-22 4 25-29 18-33 12-22 8-21


76

(b)

Figure 4.7: Impact of different number of SOP installation on power loss minimization

and voltage profile improvement


One feeder of the network from branch L25 to branch L32 as shown in Figure 4.8 was

assumed to be heavily loaded, i.e., 1.6 times higher than the loading under the normal

condition. Table 4.1 shows the impact of different number of SOPs on feeder load

balancing. It is observed that the system load balancing index LBI was reduced by

47.73% with one SOP. Although the LBI reduction was further improved with more

SOPs installed, the rate of improvement was diminishing.

Table 4.1: System Load Balancing Index with Different Number of SOP Installation

Number of SOP Installed 0 1 2 3 4

SOP Locations (Branches)

_ 25-29 25-29

18-33

25-29

18-33

12-22

26-29

18-33

12-22

8-21

System Load Balancing Index 6.156 3.218 2.566 2.428 2.389

% LBI Reduction − 47.726 58.317 60.056 61.192

0.88

0.9

0.92

0.94

0.96

0.98

1

1 4 7 10 13 16 19 22 25 28 31

Vo

ltag

e (p

.u.)

Bus Number

Impact of Number of SOP Installation on System Voltage Profile

0 SOP 1 SOP 2 SOP 3 SOP 4 SOP


77

Figure 4.8a illustrates the branch loading profile of the network. By using SOPs, the

loading of those heavily loaded branches (e.g. branches L1 to L6 and L25 to L30) was

reduced dramatically by transferring loads to the lightly loaded branches via SOP

control. As a consequence, the loading levels from branches L12 to L24 were increased.

The voltage profile was also improved, as shown in Figure 4.8b. The minimum bus

voltage (at bus 32) was 0.88 p.u. due to overloading, and it was improved by 9.09%

after the system loading was balanced using SOPs.

(a)

(b)

Figure 4.8: Impact of different number of SOP installation on load balancing and

voltage profile improvement

0

0.2

0.4

0.6

0.8

1

1.2

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Bra

nch

Lo

adin

g

Branch Number

Impact of Number of SOP Installation on Load Balancing


0.86

0.88

0.90

0.92

0.94

0.96

0.98

1.00

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33

Vo

ltag

e (p

.u.)

Bus Number

Impact of Number of SOP Installation on System Voltage Profile



78

Performance of the improved PDS method for optimal SOP

operation

Table 4.2 lists the computation time required for the calculation of above case studies

considering only one SOP. The total time required by using the conventional PDS

method (starting from an arbitrary point) and the proposed method with a good initial

approximation were compared. It can be seen that for both power loss minimization

and feeder load balancing there were significant reductions in computation time by

using the improved PDS method. Especially for solving the feeder load balancing, the

total CPU time was reduced by 68% (from 8.295s to 2.614s) after using the improved

PDS method.

Table 4.2: Results of Computing Time for Optimal SOP Operation

Optimization

technique

Power loss minimization Feeder load balancing

PDS Improved

PDS

PDS Improved

PDS

CPU time (s)

Light Loading 4.565 1.998

Overload

Condition

8.295

2.614 Normal Loading 5.364 3.343

Heavy Loading 5.774 3.578

4.5.2 Improve Network Performance Considering both SOP and

Network Reconfiguration

The benefits of combining SOP and network reconfiguration for power loss

minimization and feeder load balancing were evaluated. In this case, the SOP device

was assumed to be located between buses 25 and 29 and its power losses were ignored.

Another three case studies were also carried out for comparisons, which are

- Base case study with neither reconfiguration nor SOP;

- Case study considering only network reconfiguration;

- Case study considering only one SOP, which is located between buses 25 and 29.


79


The network performance on power loss minimization was simulated under three

loading conditions: light (50%), normal (100%), and heavy (160%). The simulation

results are listed in Table 4.3.

Table 4.3: Results of Different Methods for System Power Loss

Case Studies

Load Level

Light (0.5) Normal (1.0) Heavy

(1.6)

Base Case

No reconfiguration

No SOP installed

Switches with open status 8-21 12-22 25-29 9-15 18-33

Power Loss (kW) 47.118 202.876 575.966

Minimum Voltage (p.u.) 0.958 0.913 0.853

Only network

reconfiguration

Switches Opened 7-8 9-10 14-15 25-29 32-33

Power Loss (kW) 33.312 137.946 381.418

% Loss reduction 29.301 32.005 33.778


Only SOP

Switches with open status 8-21 12-22 9-15 18-33

SOP

Operation

𝑃𝑆_𝑖𝑛𝑗1 / MW 0.230 0.605 0.998

𝑄𝑆_𝑖𝑛𝑗1 / MVar 0.225 0.471 0.784

𝑄𝑆_𝑖𝑛𝑗2 / MVar 0.610 1.239 2.017

Power Loss (kW) 29.774 124.456 337.525

% Loss reduction 36.810 38.653 41.398


Combined

network

reconfiguration

with SOP

Switches with open status 7-8 9-10 14-15 18-33

SOP

Operation

𝑃𝑆_𝑖𝑛𝑗1 / MW 0.183 0.374 0.607

𝑄𝑆_𝑖𝑛𝑗1 / MVar 0.215 0.424 0.697

𝑄𝑆_𝑖𝑛𝑗2 / MVar 0.516 1.045 1.686

Power Loss (kW) 22.758 93.915 250.179

% Loss reduction 51.700 53.708 56.564



80

The percentages of total power loss reduction implies that using only one SOP

achieved a similar power loss reduction to that of network reconfiguration under three

loading conditions. The most significant power loss reductions and voltage

improvement under all three loading conditions were obtained using the combined

method. The SOP operation required to achieve power loss minimization indicates that

the combined method contributed more to power loss reduction while requiring smaller

SOP sizes.

Figure 4.9 illustrates the voltage profiles of all case studies under the normal loading

condition. The shapes of the voltage profile under the other two loading condition were

the same except minor change in magnitude, and hence are not illustrated.

Figure 4.9: Voltage profiles of the network under normal loading condition


The same overloading condition described in the previous case study was simulated.

Table 4.4 presents the load balancing index for the four case studies. The branch

loading profiles and voltage profiles are shown in Figure 4.10. From these results, it is

seen that by using only one SOP to control network power flows achieved better

performance on load balancing than that of using network reconfiguration to change the

0.88

0.9

0.92

0.94

0.96

0.98

1

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33

Vo

ltag

e (p

.u.)

Bus Number

Volatage Profile of 33-bus Distribution Netwo at Normal Load

Base Case Only Reconfiguration

Only SOP Combined Method


81

open/close status of a sequence of switches. The highest LBI reduction, i.e., the most

well-balanced network, and the highest improvement on minimum voltage were

achieved using the combined method.

Table 4.4: Results of Different Methods for Load Balancing

Case Studies Base

Case

Only network

reconfiguration

Only

SOP

Combined

reconfiguration with SOP

System load balancing index 6.156 4.139 3.218 2.594

LBI reduction (%) - 32.765 47.726 68.337

Minimum Voltage (p.u.) 0.880 0.904 0.928 0.943

SOP

Operation

𝑷𝑺_𝒊𝒏𝒋𝟏 / MW - - 0.916 1.142

𝑸𝑺_𝒊𝒏𝒋𝟏 / MVar - - 0.569 0.577

𝑸𝑺_𝒊𝒏𝒋𝟐 / MVar - - 1.791 2.417

(a)

0

0.3

0.6

0.9

1.2

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Bra

nch

Lo

adin

g

Branch Number

Branch Loading Using Different Methods for Load Balancing




82

(b)


improvement under different methods

4.5.3 Impact of DG Connections

A large capacity of intermittent renewable energy in the distribution network can

aggravate feeder loads unbalance and network power losses. To evaluate the benefits of

using the SOP for power loss minimization and feeder load balancing with DG

connections, three DGs were assumed to be connected to buses 16, 17 and 18. Each DG

had a power production of 1 MW with a unity power factor and was modelled as a

negative load in this study. An SOP was assumed to be located between buses 18 and

33 and its power losses were ignored. Five case studies were carried out for

comparisons, which are

− Base case study with and without DG connections;

− Case study using only network reconfiguration;

− Case study using only one SOP;

− Case study using the combined network reconfiguration and SOP.

0.87

0.90

0.93

0.96

0.99

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33

Vo

ltag

e (p

.u.)

Bus Number

Volatage Profile of 33-bus Distribution Network




83


Table 4.5 shows the network power losses for the five case studies under normal

loading conditions. It can be seen that the network power losses were increased by 70%

due to the DG connections. A significant power loss reduction (72%) was achieved by

using only one SOP, which gave better performance than by using only network

reconfiguration. The most significant power loss reduction was obtained by using the

combined method while requiring smaller SOP size. Figure 4.11 shows the voltage

profile of the network. There was a notable voltage rise due to the DG connections. The

use of SOP (either with or without network reconfiguration) achieved much flatter

network voltage profiles.

Table 4.5: Results of Different Methods for Power Loss Minimization with DG

Connections

Case Studies Base Case

without DGs

Base Case

with DGs

Only Network

Reconfiguration

Only

SOP

Combined

Method

Total Power Losses (kW) 202.875 345.502 120.783 92.334 63.221

% Loss Reduction - - 65.044 73.275 81.702

SOP

Operation

𝑃𝑆_𝑖𝑛𝑗1 / MW - - - 1.661 1.606

𝑄𝑆_𝑖𝑛𝑗1 / MVar - - - 0.376 0.381

𝑄𝑆_𝑖𝑛𝑗2 / MVar - - - 0.899 0.904

Figure 4.11: Voltage profiles of the network under different cases with DG connection

0.94

0.96

0.98

1.00

1.02

1.04

1.06

1.08

1.10

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33

Vo

ltag

e (p

.u.)

Bus Number

Volatage Profile of 33-bus Distribution Network with DG connections

Base Case with DG connection Only reconfiguration



84


To evaluate the benefits of the SOP for feeder load balancing with DG connections,

(especially when one feeder has high DG power generation while the other one is under

a heavy loading condition,) the same overloading condition adopted previously was

used again. Table 4.6 shows the load balancing index for the five case studies. It is

observed that the network was more unbalanced (higher LBI) after connecting DGs. A

significant reduction on LBI (77.51%) was achieved by transferring loads from the

heavily loaded feeder to that with DG integration via the SOP, which gave much better

performance than network reconfiguration. The most well-balanced network was

achieved by using the combined method with smaller SOP size required.

Table 4.6: Results of Different Methods for Load Balancing with DG Connections

Case Studies Base Case

without DGs

Base Case

with DGs

Only Network

Reconfiguration

Only

SOP

Combined

Method

System Load Balancing Index 6.156 7.238 2.995 1.628 1.267

% LBI Reduction - - 58.621 77.507 82.495

SOP

Operation

𝑃𝑆_𝑖𝑛𝑗1 / MW - - 1.692 1.480 1.606

𝑄𝑆_𝑖𝑛𝑗1 / MVar - - 0.457 0.523 0.381

𝑄𝑆_𝑖𝑛𝑗2 / MVar - - 1.252 1.211 0.904

Figure 4.12 shows the branch loading profiles and voltage profiles. As shown in Figure

4.12a, the peak branch loading of the network was reduced from 72% to 50% by using

only one SOP. It shows that the increase in peak currents in the feeders and branch

loading was reduced effectively by using only one SOP. The combined method

achieved the lowest peak branch loading. The voltage profile was also much flatter by

using the SOP and the combined method, as shown in Figure 4.12b.


85

(a)

(b)

Figure 4.12 Results of load balancing capability and relevant voltage profile improvement

under different cases with DG connections

4.5.4 Impact of the SOP Device Losses

A significant reduction on feeder power losses using SOPs has been demonstrated in

the previous study, where the SOP device losses were not considered. Although the

SOP device losses have minor impact on its power flow control, it may lower the

economic benefits obtained from reducing the total network power losses. The impact

of SOP device losses on the total network power loss reduction was investigated based

on sensitivity analysis. Different VSC efficiencies at the rated power, i.e. the values of

loss coefficients kI and kJ in (4-11), were considered. For the VSCs in the back-to-back

converters, it was assumed that kI is equal to kJ. The constant power loss of the SOP

0

0.2

0.4

0.6

0.8

1

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Bra

nch

Lo

adin

g

Branch Number

Branch Loading Using Different Methods for Load Balancing with DG connections

Base Case with DGs Only Reconfiguration


0.90

0.92

0.94

0.96

0.98

1.00

1.02

1.04

1.06

1.08

1.10

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33

Vo

ltag

e (p

.u.)

Bus Number

Volatage Profile of 33-bus Distribution Network with DG connections

Base Case with DGs Only reconfiguration



86

device was set as 0.2% of the rated power.

Figure 4.13a illustrates the total network power loss minimization (feeder losses and

SOP device losses) under the normal loading conditions. One SOP, located between

buses 25 and 29 (without DG connections), was considered in this case. Different

device efficiencies of each VSC were considered from 93.5% to 100%. It is observed

that the total network losses were reduced by using the SOP. However, the benefits

were lowered when the device efficiency decreased. The SOP lost its capability in

reducing network power losses when the device loss coefficients fell to 0.065, i.e., 93.5%

VSC efficiency at the rated power. Figure 4.13b shows the performance on total

network power losses for each percentage of system loading from 0 to 60%. The figure

shows that SOP had a greater positive impact on the network power losses under higher

system loading conditions. The requirements on SOP device efficiency for total

network loss reduction were reduced when the system loading increases.

(a)

0

25

50

75

100

125

150

175

200

Base Case 0 0.01 0.015 0.02 0.025 0.065

Tota

l Po

wer

Lo

sses

(kW

)

Device Loss Coefficient kI /kJ

SOP Power Losses Feeder Power Losses


87

(b)

Figure 4.13: Impacts of the SOP device losses on total network power losses

4.6 Summary

The benefits of using SOPs for medium voltage distribution networks were investigated

focusing on power loss reduction, feeder load balancing and voltage profile

improvement. A generic power injection model of SOP that is suitable for steady state

analysis was developed, taking into account both physical limits and internal power

losses of a typical SOP device: back to back VSCs. The optimal SOP operation is

obtained using improved Powell’s Direction Set method. The combined method

considering both SOP and network reconfiguration was proposed to demonstrate the

superiority of using SOPs.

Different number of SOPs was considered. Results showed that SOPs significantly

contributed to power loss reduction, feeder load balancing and voltage profile

improvement. By comparing with network reconfiguration, using only one SOP

achieved similar improvement on network power loss reduction and feeder load

balancing. The greatest improvements were obtained when combining SOP and

network reconfiguration where smaller SOP sizes were required.

100

200

300

400

500

600

0 10 20 30 40 50 60

Tota

l Po

wer

Lo

lsse

s (k

W)

Percentage of Load Increase (%)

Power Losses in Base Case

No Device Losses

99% VSC Efficiency at Rated Power

98% VSC Efficiency at Rated Power


88

The impact of the penetration of DGs was investigated. Results showed that SOPs are

able to significantly reduce the peak currents in feeders and alleviate undesirable

voltage excursions induced by the connection of DG and demand. Therefore SOPs can

be used as an alternative to infrastructure upgrades in accommodating distributed

energy resources.

The impact of SOP device losses was illustrated which shows that the economic benefit

of SOP in reducing total network power losses was lowered with a decrease of device

efficiency. However, a greater positive impact was obtained when system loading

increased. In other words, the requirements on SOP device efficiency for total network

loss reduction were reduced when the system loading increases.

Chapter 5 Voltage Control in Active Distribution Networks with SOPs

89

Voltage Control in Active

Distribution Networks with

SOPs

HIS chapter reports a novel coordinated voltage control strategy for

active distribution networks considering the SOP, the OLTC and

multiple DG units as the voltage control devices.

Chapter 5

T


90

5.1 Introduction

Chapter 3 illuminated the excellent performance of an SOP in controlling real and

reactive power flow under normal operating conditions using back-to-back VSCs.

Although it is anticipated that SOPs will support distribution network voltage control

by changing the real and reactive power flow of the network, the control strategy

regarding the exact operation of the SOP and the coordination with other voltage

control devices for distribution networks needs to be investigated.

In this chapter, a novel coordinated voltage control strategy was proposed considering

the controllable devices of the OLTC, the SOP and multiple DG units. The performance

of the control strategy (using single device or coordination among multiple devices)

were assessed considering the voltage compliance, number of tap operations, daily DG

energy integration and network energy losses as the key metrics. A distributed control

framework was used, where each of the controllable elements is considered as a local

control agent and operates based on the local measurements and the shared data

through an information-sharing platform. This coordination was priority-based, where

the local control agents, without losing the independence of each, can achieve an

optimal voltage management with few information exchanges. In the case study of this

work, the proposed control strategy was applied to an 11-kV example network

considering different network operating conditions and DG penetrations.

5.2 Voltage Control Framework

In medium voltage (MV) networks, voltage control devices usually make control

actions based on the adequate knowledge of the status of the network. Due to the

limited visibility in MV networks, it is desirable to achieve a better voltage

management based on some estimations of the voltage profile using the existing

real-time measurements and advanced information and communication devices.


91

5.2.1 Voltage Profile Estimation

In an MV network, the voltage control actions can be taken base on the maximum and

minimum voltages in the network. Figure 5.1 shows the voltage profiles of an MV

feeder with and without DG. For a passive network without DG connections (assuming

no SOP connection in this case), the voltage profile along the feeder decreases, as

shown in Figure 5.1 (profile a). Therefore, the maximum and minimum voltages are

located at the substation busbar (i.e., the secondary of the transformer) and the remote

end of the feeder, respectively. For an active network with DG units, the voltage in the

feeder can be higher than the substation busbar, as shown in Figure 5.1 (profile b).

Moreover, SOPs is able to change the network power flow and thus change the voltage

profile. Therefore, for a network with DG connections and/or SOPs, the maximum

voltage can locate either at the substation busbar, the DG or SOP connecting buses

whilst the minimum voltage point can locate at any bus including in and between the

substation busbar, DG or SOP connecting bus.

V0 1 2 3 4 5

DG1 DG2

V0

V

Voltage profile

a

b

bus

Load

Figure 5.1: Voltage profiles of an MV feeder with and without DG

In this work, it is assumed that the maximum voltage of an MV feeder is obtained from


92

the meter readings at the substation busbar, DG and SOP connecting buses. A

simplified estimation for the minimum voltage is carried out based on these meter

readings and necessary calculations.

Figure 5.2 shows an MV feeder with several DG units and an SOP connected. Meters

are placed at the substation busbar, the SOP point and each DG connecting bus. The

feeder line between each two of them is considered as a segment.

DG DG

SOP

Segment

Meters

Figure 5.2: An MV feeder with segments for the minimum voltage estimation

The minimum voltage for each of these segments is first estimated by using the meter

readings of the corresponding two terminals and necessary calculations (shown below),

and then the minimum for the whole feeder is obtained by comparing all these values.

Mathematically, the estimation of the minimum voltage within a segment can be

implemented through two steps:

1) Checking the existence of a minimum voltage point within a segment

2) Estimating the value of the minimum voltage if it exists.

Details of these two steps are presented as follows.

Step1) Checking the existence of a minimum voltage point within a

segment

It is observed from Figure 5.1 that there is a minimum voltage point in between the two


93

DG connecting buses if and only if, for both DGs, the voltage of the DG connecting

bus is higher than the voltage of its neighbouring bus (in the direction of the other DG).

For instance, in Figure 5.1, there will be a minimum voltage point in between bus 1 and

bus 5 if and only if the voltage of bus 1 is higher than the voltage of bus 2 while the

voltage of bus 5 is higher than the voltage of bus 4. Equation (5-1) shows the

mathematical condition for the existence of a minimum voltage point between DG 1

connecting bus and DG 2 connecting bus:

{𝑉1 − 𝑉2 > 0𝑉5 − 𝑉4 > 0

(5-1)

To solve the above inequality equations, the power flow in the feeder, as shown in

Figure 5.3, is calculated by the following set of recursive equations [93]:

𝑃𝑖 = 𝑃𝑖−1 − 𝑃𝐿𝑜𝑠𝑠 (𝑖−1, 𝑖) − 𝑃𝐿,𝑖 = 𝑃𝑖−1 −𝑟𝑖−1

𝑉𝑖−12 ∙ (𝑃𝑖−1

2 + 𝑄𝑖−12) − 𝑃𝐿,i (5-2)

𝑄𝑖 = 𝑄𝑖−1 − 𝑄𝐿𝑜𝑠𝑠 (𝑖−1, 𝑖) − 𝑄𝐿,𝑖 = 𝑄𝑖−1 −𝑥𝑖−1

𝑉𝑖−12 ∙ (𝑃𝑖−1

2 + 𝑄𝑖−12) − 𝑄𝐿,i (5-3)

𝑉𝑖2 = 𝑉𝑖−1

2 − 2 ∙ (𝑟𝑖−1𝑃𝑖−1 + 𝑥𝑖𝑄𝑖) +(𝑟𝑖−1

2+𝑥𝑖−12)

𝑉𝑖−12 ∙ (𝑃𝑖−1

2 + 𝑄𝑖−12) (5-4)

where 𝑃 represents active power, 𝑄 reactive power, 𝑉 nodal voltage, 𝑟 resistance

and 𝑥 reactance. Subscript 𝐿𝑜𝑠𝑠 denotes line losses, and 𝐿 denotes load.

Vi-1 ViV0

ri-1+xi-1 j

Pi-1+Qi-1 j

PL,i+QL,i j

Pi+Qi j

PL,i-1+QL,i-1 j

Figure 5.3: A part of distribution feeder

Ignoring the quadratic terms (line losses) since it is much smaller than the branch

power [99], equation (5-2)-(5-4) can be written as


94

𝑃𝑖 = 𝑃𝑖−1 − 𝑃𝐿,i (5-5)

𝑄𝑖 = 𝑄𝑖−1 − 𝑄𝐿,i (5-6)

𝑉𝑖2 = 𝑉𝑖−1

2 − 2 ∙ (𝑟𝑖−1𝑃𝑖−1 + 𝑥𝑖𝑄𝑖) (5-7)

Based on (5-7), (5-1) can be represented as follows:

{𝑉1 − 𝑉2 =

2∙(𝑃1∙𝑟1+𝑄1∙𝑥1)

𝑉1+𝑉2> 0

𝑉4 − 𝑉5 = −2∙(𝑃4∙𝑟4+𝑄4∙𝑥4)

𝑉4+𝑉5> 0

(5-8)

After further simplification on (5-8), the condition for the existence of a minimum

voltage point between DG 1 and DG 2 can be given as

{ 𝑃1 ∙ 𝑟1 + 𝑄1 ∙ 𝑥1 > 0𝑃4 ∙ 𝑟4 + 𝑄4 ∙ 𝑥4 < 0

(5-9)

where the reactive and reactive power are measured locally by using meters at the DG

locations. Note that (5-9) can be used as a generic condition for checking the existence

of a minimum point within any segment, such as the segment between the substation

busbar and one DG, one DG and one SOP, etc.

Step 2) Estimating the value of the minimum voltage within a segment

A simplified estimation method is used to obtain the value of the minimum voltage if it

exists. This method gives a worst case value thus it is considered as a good lower

bound for the voltage management purpose.

The loads in between the two DG connecting buses in Figure 5.1 are assumed to be

concentrated and connected half way between the two DG connecting buses, as shown

in Figure 5.4.


95

V1

r/2+x/2 j

P1+Q1 j

DG2DG1

V5Vmin_estP4+Q4 j

r/2+x/2 j

PL, min = P1-P4

QL, min = Q1-Q4

Figure 5.4: Voltage estimation using simplified method

Therefore, the minimum voltage between the DG 1 and DG 2 calculated from DG 1

using (5-7) is

𝑉min _𝑒𝑠𝑡,𝐷𝐺1 = √𝑉12 − 2 ∙ (𝑃1 ∙

𝑟

2+ 𝑄1 ∙

𝑥

2) (5-10)

Similarly, the minimum voltage calculated from DG 2 is

𝑉min _𝑒𝑠𝑡,𝐷𝐺2 = √𝑉52 + 2 ∙ (𝑃4 ∙

𝑟

2+ 𝑄4 ∙

𝑥

2) (5-11)

where r and x are the total value of the reactance and resistance between DG 1 and DG

2.

Taking the minimum value of 𝑉min _𝑒𝑠𝑡,𝐷𝐺1 and 𝑉min _𝑒𝑠𝑡,𝐷𝐺2 as the minimum voltage

of the segment, 𝑉𝑚𝑖𝑛, as shown in (5-12).

𝑉min _𝑒𝑠𝑡 = 𝑚𝑖𝑛 {𝑉min _𝑒𝑠𝑡,𝐷𝐺1, 𝑉min _𝑒𝑠𝑡,𝐷𝐺2} (5-12)

Equation (5-12) provides an estimation of the minimum voltage using the data

measured at the substation, DG or SOP connecting buses.

Note that above calculation used to estimate the minimum voltage of an MV feeder did

not consider feeders having long laterals or DGs connected in the lateral. In this work,

the loads and DGs connected to the laterals are assumed to be lumped into the nearest

main stream of the feeder and hence errors exist.


96

5.2.2 Proposed Voltage Control Framework

A distributed voltage control framework is proposed, as depicted in Figure 5.5. Each of

the controllable elements, i.e. the OLTC, the SOP and DG units, is considered as a local

control agent that operates based on the local measurements and the shared data

through an information-sharing platform.

DG

DG DG SOPOLTC

OLTC

AgentDG

Agent

SOP

Agent

Information-sharing Platform

Communication links

Meters

Figure 5.5: Distributed voltage control framework

Local control agents

The local control agents carry out real-time control based on the local measurements

but considering coordination with other local control agents via the information

exchange through the information-sharing platform. Figure 5.6 shows the interior

structure of each control agent.


97

Voltage estimation

module

Control algorithm

module

Knowledge base

Bidirectional interface

Local meters Actuators

Information-

sharing platform

Action

Figure 5.6: Interior structure of a local control agent

Details of the four blue blocks shown in Figure 5.6 are given as follows.

Bidirectional interface is responsible for 1) exchanging data with the

information-sharing platform, 2) passing the measurements from the local meters and

the information-sharing platform to the voltage estimation module inside the agent, and

3) passing the control commands issued by the control algorithm module to the

actuator.

Voltage estimation module is responsible for performing the segment voltage

estimation as described in 5.2.1. The detailed process is shown as follows.

1) The module receives the measurements from the local meter. Figure 5.7 shows a

detailed view of the measurements by each local meter. It includes the voltage

magnitude at the connecting bus, the real and reactive power flow in lines

connected to the connecting bus. These measurements provide the measured

potential maximum voltage and the data needed to estimate the minimum

voltages within its corresponding segments, i.e., segments in between its agent

and any neighbour agent.


98

V

P, Q P, Q

Local Meter

segmentsegment

Figure 5.7: Details of local measurements

2) According to Step 1 in Section 5.2.1, the module checks one part of the condition

for the existence of a minimum voltage point within a segment.

3) If the minimum voltage point is found exist, the module estimates the value

according to Step 2 in Section 5.2.1. Otherwise, goes to 4).

4) The module sends voltage data to the bidirectional interface and finally to the

information-sharing platform. Noted that only the critical data are sent to the

information-sharing platform, i.e., the measured node voltage and estimated

minimum voltages within the corresponding segments.

Control algorithm module provides local control commends using the corresponding

control algorithm.

Knowledge base stores the knowledge information of the network, e.g. network

topology, line parameters.

Information-sharing platform

The Information-sharing platform works as a moderator: 1) determining the overall

maximum and minimum voltages of each feeder and passing onto the local control

agent if needed; and 2) perform the coordinated voltage control strategy, which are

described in Section 5.3.


99

5.3 Coordinated Voltage Control Strategy

In this section, the voltage control using the SOP and its coordination with other control

devices (i.e., OLTC and DG units) were investigated. The effectiveness of the SOP on

voltage control and its impacts on networks was first examined. Then, a priority-based

coordination using the proposed control framework described in Section 5.2 was

developed.

5.3.1 Effectiveness of SOP on Voltage Control

Voltage change caused by an SOP

Figure 5.8 shows a radial feeder of an MV network.

V0 Vi-1 Vi Vn

PL,i+QL,i j

ri+xi j

Vi+1

ri-1+xi-1 j

Pi+Qi jPi-1+Qi-1 j Pn+Qn j

Figure 5.8: Radial distribution feeder

According to (5-5)-(5-7), the active and reactive power at node i and the voltage

magnitude at node i can be given as follows:

𝑃𝑖 = 𝑃𝑛 + ∑ 𝑃𝐿,𝑘𝑛𝑘=𝑖+1 (5-13)

𝑄𝑖 = 𝑄𝑛 + ∑ 𝑄𝐿,𝑘𝑛𝑘=𝑖+1 (5-14)

𝑉𝑖2 = 𝑉0

2 − 2 ∙ ∑ (𝑟𝑘𝑃𝑘 + 𝑥𝑘𝑄𝑘)𝑖−1𝑘=0 (5-15)

Equations (5-13) to (5-15) show the change of the branch power flow and consequently

the change of the voltage magnitude at bus i depends on the sum of power loads

(𝑃𝐿,𝑘 and 𝑄𝐿,𝑘), the downstream active and reactive power flow (𝑃𝑛 and 𝑄𝑛) and the


100

voltage magnitude at the initial upstream bus (𝑉0), i.e. the substation busbar.

Therefore, if an SOP is connected in the downstream of bus i. According to the generic

power injection model of SOP developed in Chapter 3, the change of the branch power

at bus i caused by the change of the power injection，∆𝑃𝑆_𝑖𝑛𝑗 and ∆𝑄𝑆_𝑖𝑛𝑗 at the SOP

connecting bus can be estimated based on (5-13) and (5-14):

∆𝑃𝑖 = 𝑃𝑖𝑛𝑒𝑤 − 𝑃𝑖

𝑜𝑙𝑑 = ∆𝑃𝑛 = ∆𝑃𝑆_𝑖𝑛𝑗 (5-16)

∆𝑄𝑖 = 𝑄𝑖𝑛𝑒𝑤 − 𝑄𝑖

𝑜𝑙𝑑 = ∆𝑄𝑛 = ∆𝑄𝑆_𝑖𝑛𝑗 (5-17)

Thus, the estimation on the change of the voltage magnitude at bus i using (5-15) can

be obtained as

∆(𝑉𝑖2) = 𝑉𝑖

𝑛𝑒𝑤2−𝑉𝑖

𝑜𝑙𝑑2= (𝑉0

𝑛𝑒𝑤2−𝑉0

𝑜𝑙𝑑2) − 2 ∙ ∑ 𝑟𝑘 ∙ ∆𝑃𝑆𝑖𝑛𝑗

𝑖−1𝑘=0 − 2 ∙ ∑ 𝑥𝑘 ∙ ∆𝑄𝑆𝑖𝑛𝑗

𝑖−1𝑘=0

= ∆(𝑉02) − 2 ∙ ∑ 𝑟𝑘 ∙ ∆𝑃𝑆𝑖𝑛𝑗

𝑖−1𝑘=0 − 2 ∙ ∑ 𝑥𝑘 ∙ ∆𝑄𝑆𝑖𝑛𝑗

𝑖−1𝑘=0 (5-18)

where ∆(𝑉02) is the change of the substation voltage; 2 ∙ ∑ 𝑟𝑘

𝑖−1𝑘=0 and 2 ∙ ∑ 𝑥𝑘

𝑖−1𝑘=0

are the sensitivity factors of bus i. These sensitivity factors constants determined by the

line parameters only.

While, for bus i in the downstream of SOP connecting bus, the change of the voltage at

bus i equals to the change of the voltage at the SOP connecting bus j, which is written

as follows:

∆(𝑉𝑖2) = ∆(𝑉0

2) − 2 ∙ ∑ 𝑟𝑘 ∙ ∆𝑃𝑆_𝑖𝑛𝑗𝑗𝑘=0 − 2 ∙ ∑ 𝑥𝑘 ∙ ∆𝑄𝑆_𝑖𝑛𝑗

𝑗𝑘=0 (5-19)

Note that the change of the substation voltage, i.e., ∆(𝑉02) is calculated based on the

transformer impedance and the system input impedance at the primary side of the


101

substation. It is assumed that this change is small and thus ignored in this study.

Equations (5-18) and (5-19) also applies to the DG unit for voltage control in this work.

The impacts of the SOP on MV networks for voltage control

According to the UK statutory instruments, the Distribution Network Operator (DNO)

is obligated to maintain the voltages on MV networks within ± 6% of nominal [100]. In

this work, voltage limits of ± 5% of nominal, i.e. 1.05 and 0.95 p.u. were considered so

as to leave some headroom for unexpected voltage violation due to the errors of

measurements, estimations and communication delays.

An SOP controls network voltage by either supplying or absorbing reactive power at its

connected feeders or controlling the active power flow between the feeders. According

to (5-18) and (5-19), using either reactive or active control can achieve a desired

voltage at the targeting bus. The corresponding impacts on network voltage profile and

network operation (line losses) was tested in a simple MV network shown in Figure 5.9.

The voltage operating limits of this test network was considered ±5% of nominal. A

2.5 MW DG with unity power factor was assumed to be connected to bus 13 in feeder 1.

The network parameters can be found in appendix A.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

01 2 3 4 5 6 7 8

HV/MV

Transformer

DG

SOP

Feeder 2

Feeder 1

Figure 5.9: Two-feeder MV distribution network

Figure 5.10 shows the voltage profile of the two feeders before voltage control. The


102

voltages from bus 12 down to bus 17 violated the upper limit due to the DG connection,

while the minimum voltage (on the other feeder) approached to the lower limit. The

OLTC is infeasible for voltage control in such a case since the difference between the

overall maximum and minimum voltages exceeded the gap between the upper and

lower voltage limits and thus may cause improper settings and excessive tap operation

[101]. Therefore, the SOP was used to reduce the voltages in feeder 1 by controlling

either real or reactive power based on equation (5-18).

Figure 5.10: Voltage profiles of both feeder before voltage control

According to (5-18), it was assumed that the target bus is 13 and the desired voltage

magnitude at the target bus, 𝑉13𝑛𝑒𝑤 was set to be 1 p.u. Figure 5.11 shows the voltage

profile with SOP control using different control methods. It is observed that both

control methods achieved the control task, i.e., the voltages at bus 13 were close to the

targeted value (1 p.u.).For active power control shown in Figure 5.11a, the voltage

profile along feeder 2 rose while the voltage profile along feeder 1 dropped. This is due

to the active power was transferred from feeder 1 to feeder 2. However, for the reactive

control method, the reactive power was controlled independently on either side of the

SOP and thus the voltage profile along the feeder 2 was not changed. In addition, it is

shown that the closer to the SOP connecting bus, the bigger the increased (or decreased)

voltage magnitude.. This can be explained by (5-19) where the sensitivity factors, 2 ∙

0.95

0.975

1

1.025

1.05

1.075

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Volt

age

(p.u

.)

Bus Number

Feeder 1 Feeder 2

Upper limit

Lower limit


103

∑ 𝑟𝑘𝑖−1𝑘=0 and 2 ∙ ∑ 𝑥𝑘

𝑖−1𝑘=0 increase with the distance between bus i and the substation

bus increases.

(a)

(b)

Figure 5.11: Voltage profiles with SOP voltage control: (a) active power control method;

(b) reactive power control method

Figure 5.12 compares the network power losses before and after using SOP control.

There was a significant increase of network power losses after using the reactive power

control method. This is because the voltage improvement was achieved by absorbing

excessive reactive power from the network, which caused an increased in feeder

loading and consequently an increase in the line losses. On the contrary, the network

power losses were decreased by using the active power control method which

transferred active power between feeder 1 and feeder 2.

0.95

0.975

1

1.025

1.05

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Vo

ltag

e (p

.u.)

Bus Number

SOP_P control - feeder 1

SOP_P control - feeder 2

0.925

0.95

0.975

1

1.025

1.05

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Volt

age

(p.u

.)

Bus Number

SOP_Q control - feeder 1

SOP_Q control - feeder 2


104

Figure 5.12: Comparison of the network active power losses

Discussion

According to the above analysis, the following key findings of the SOP control are

outlined:

Either real or reactive power control of the SOP achieves the voltage control

objective by using the sensitivity factors in (5-18). However, the voltage change

is more sensitive to the active power control due to the fact that MV distribution

networks have high line resistances and low X/R ratio [102]. This makes the

active power control method more cost-effective considering the device capacity.

For active power control method, the voltage profile on both feeders will be

affected. A decrease in the voltage profile on one feeder leads to an increase in

the voltage on the other feeder, and vice versa. Therefore, the voltage headroom

of both connected feeders need to be considered when the active power control

method is used. The main advantage of this control method compared with the

reactive power control is that there will not be a significant change on network

power losses using the active power control.

The main advantage of the reactive power control method is that both sides of

SOP supply and/or absorb reactive power independently. Therefore, it enables

voltage control simultaneously and independently on both connected feeders.

However, the voltage drop by absorbing excessive reactive power may cause a

165.1452

101.158

434.5934

0

100

200

300

400

500

Original SOP_P control SOP_Q control

To

tal

po

wer

lo

ss (

kW

)


105

significant increase in network power losses.

5.3.2 Priority-Based Coordination

Figure 5.13 shows a priority-based coordination among multiple agents, including the

OLTC, the SOP and the DG units. Each of these agents is called a “local control agent”

and each local control agent has multiple objectives. The first one (main objective) is

identical to all agents and the remaining are called sub-objectives. The main objective

used in this work is to keep the voltages within the allowable range. The sub-objectives

are to reduce the number of tap operations, mitigate DG active power curtailment, and

reduce network power losses.

OLTC

Agent

DG

Agent

SOP

Agent

Information-sharing

Platform

No voltage violations

Less active power curtailment

Fewer network power losses


Fewer number of tap operations



Exchange

messages

Exchange

messages

Exchange

messages

Figure 5.13: Priority-based coordination of the multiple local control agents

In order to achieve a trade-off among the aforementioned objectives, the objectives

were grouped by defining the priority level, as shown in Figure 5.14.


106

No voltage violation

Less tap operation


Maximize energy capture

Priority LevelH

L

Figure 5.14: Prioritisation of the proposed objectives from high (H) to low (L)

A priority-based coordination is carried out to allow the local control agents to achieve

their objectives in accordance with a specific sequence (i.e., the prioritisation). Figure

5.15 shows the flow chart of the priority-based coordination. Both DG and SOP control

agents have two voltage control methods: real and reactive power control methods. The

OLTC agent also has two control strategies to deal with both feasible and infeasible

solutions. The main objective (no voltage violation) is considered each time before a

control method to determine whether a control action is needed. All the control

methods are coordinated in a pre-defined sequence based on the prioritisation of the

sub-objectives. The higher the priority level, the later the corresponding control method

will be activated.

Thus, as shown in the figure 5.15, the real and reactive power control of the SOP and

the reactive control of the DG (green block) are the first considered methods. They are

activated in a sequence that: 1) the active power control of the SOP agent is prior to the

reactive power control methods due to the results obtained from Section 5.3.1, i.e., the

active power control method is more sensitive to voltage change and results in reduced

network power losses and 2) the reactive control method of the SOP agent is prior to

that of the DG agent in order to allow more flexibility for the DG unit in case of an

emergency [56]. All these real and reactive control methods are performed by solving

an optimization problem to minimize the voltage deviation, and thus, reduce the

number of tap operations. The second activated method is the OLTC control (yellow

block). A double-check process (grey block) is introduced before activating the last

control method in order to ensure that there is no feasible solutions to enable an OLTC


107

control (feasible solution strategy). The last activated control method is the active

power curtailment by the DG, as the brown block shown.

Start

SOP_Real power control method

Voltage violation?

SOP_Reactive power control method

Voltage violation?

DG_Reactive power control method

Feasible solution?

Stop double check ?

OLTC_Feasible

solution strategyEnd

OLTC_Infeasible solution strategy

DG_Real power curtailment

Voltage violation?

Yes

Yes

Yes

Yes

Yes No

No

No

No

Less network power losses

Less tap operation

Maximize energy capture

Double check processNo

Voltage violation?No

Yes

Double check

process

Voltage violation?No

Yes

Figure 5.15: Flow chart of the proposed priority-based coordination

The details of the voltage control method for each agent and the double check process

are given as follows.


108

Voltage control for the SOP agent

A. Bidirectional interface

The SOP control agent receives data from the information-sharing platform. These data

contains the action command, magnitudes and locations of all potential maximum and

minimum voltages along the connected feeders. The action command specifies the

method to be activated at current time step. The locations of all potential maximum and

minimum voltage points are sent to the knowledge based to calculate the sensitivity

factors.

After the control algorithm module determines the required real or reactive power for

the SOP agent, two messages are sent out via the interface: the new real or reactive

power set point to the actuator and the action confirmed message to the

information-sharing platform.

B. Control algorithm module

The control algorithm module implements the control algorithm based on the action

command, voltage data and the corresponding sensitivity factors.

1) Active power control algorithm

The active power control algorithm solves the optimization problem shown in

(5-20)-(5-23). The objective function aims to minimize the voltage deviation of the

maximum and minimum voltages along both connected feeders from the nominal value,

𝑉𝑛𝑜𝑟𝑚:

Minimize 𝐹(𝑃𝑆_𝑖𝑛𝑗𝑛𝑒𝑤 ) = ∑ (𝑉𝑖 − 𝑉𝑛𝑜𝑟𝑚)2𝑁

𝑖=1 (5-20)

subject to

|𝑃𝑆_𝑖𝑛𝑗𝑛𝑒𝑤 | ≤ 𝑃𝑆_𝑖𝑛𝑗

𝑙𝑖𝑚 , 𝑃𝑆_𝑖𝑛𝑗𝑙𝑖𝑚 = √𝑆𝑆𝑂𝑃

𝑟𝑎𝑡𝑒2− 𝑄𝑆_𝑖𝑛𝑗

2 (5-21)


109

𝑉𝑖𝑛𝑒𝑤2

= 𝑉𝑖𝑜𝑙𝑑2

− 𝑀𝑖 ∙ (𝑃𝑆_𝑖𝑛𝑗𝑛𝑒𝑤 − 𝑃𝑆_𝑖𝑛𝑗

𝑜𝑙𝑑 ) (5-22)

𝑉𝑚𝑖𝑛𝑏𝑎𝑛𝑑 ≤ 𝑉𝑖 ≤ 𝑉𝑚𝑎𝑥

𝑏𝑎𝑛𝑑 (5-23)

where 𝑉𝑖 is the voltage magnitude received from the interface; N is the total number of

voltage points received from the interface; 𝑆𝑆𝑂𝑃𝑙𝑖𝑚 is the rating power of the SOP device;

𝑀𝑖 is the sensitivity factor for the change of voltage 𝑉𝑖 with respect to the change of

active power; 𝑉𝑚𝑖𝑛𝑏𝑎𝑛𝑑 and 𝑉𝑚𝑎𝑥

𝑏𝑎𝑛𝑑 are the acceptable limits of bandwidth for voltage

control using the SOP agent. In this work, 𝑉𝑛𝑜𝑟𝑚 is assumed to be 1 p.u., i.e., a middle

point between the upper and lower limits in order to leave headroom for both over- and

under-voltage issues due to the future DG penetrations and increase of demands, e.g.,

electric vehicles and heat pumps.

Note that the defined bandwidth in (5-23) is wider than the specified limits to avoid the

infeasible solution, which may occur due to the device power limits in (5-21). This

wider bandwidth is also applied to the reactive power method of the SOP and DG agent

and is refined by using the OLTC agent to ensure that the network voltages within the

statutory limits. The optimization algorithm used in this study was the Powell’s Direct

Set Method, which is described in Chapter 4.

2) Reactive power control algorithm

The aim of reactive power control is to minimize the voltage deviation from the

nominal value and the amount of reactive power absorbed by the SOP to improve

voltage control. The objective function is defined as

Minimize 𝐽 (𝑄𝑆𝑖𝑛𝑗

𝑛𝑒𝑤) = ∑ (𝑉𝑖 − 𝑉𝑛𝑜𝑟𝑚)2𝑁𝑖=1 − 𝜔𝑎𝑏𝑠𝑜𝑟 ∙ 𝑚𝑖𝑛{0, 𝑄𝑆𝑖𝑛𝑗

𝑛𝑒𝑤} (5-24)

subject to


110

|𝑄𝑆𝑖𝑛𝑗

𝑛𝑒𝑤| ≤ 𝑄𝑆_𝑖𝑛𝑗𝑙𝑖𝑚 , 𝑄𝑆_𝑖𝑛𝑗

𝑙𝑖𝑚 = √𝑆𝑆𝑂𝑃𝑙𝑖𝑚 2

− 𝑃𝑆_𝑖𝑛𝑗2 (5-25)



− 𝐾𝑖 ∙ (𝑄𝑆_𝑖𝑛𝑗𝑛𝑒𝑤 − 𝑄𝑆_𝑖𝑛𝑗

𝑜𝑙𝑑 ) (5-26)


𝑏𝑎𝑛𝑑 (5-27)

where 𝑄𝑆𝑖𝑛𝑗

𝑛𝑒𝑤 is the reactive power supplied by the SOP, and it turns to be a negative

value when the SOP absorbs reactive power from the connected feeder; 𝜔𝑎𝑏𝑠𝑜𝑟 is the

penalty factor for absorbing reactive power from the connected feeder in order to avoid

the increase in the losses along the feeder, i.e., the term −𝜔𝑎𝑏𝑠𝑜𝑟 ∙ 𝑚𝑖𝑛{0, 𝑄𝑆𝑖𝑛𝑗

𝑛𝑒𝑤} will

turn to be a non-zero (positive) value and add to the objective function when the SOP

absorbs reactive power from the network; 𝐾𝑖 is the sensitivity factor for the change of

voltage 𝑉𝑖 with respect to the change of reactive power.

Voltage control for the DG agent

In this work, the DG units are assumed to operate at the maximum power point tracking

(MPPT) before receiving the voltage control command from the information-sharing

platform. Once receiving a command, the DG agents will accept or refuse the command

based on the assessment of the request, for instance, assessing the remaining capacity

for voltage support.


The DG agent receives data from the information-sharing platform. Similar to the SOP

agent, these data contains the action command, magnitudes and locations of potential

maximum and minimum voltages along the connected feeder. The action command

specifies the control method to be activated at current time step, i.e., reactive power

control or active power curtailment. The locations of potential maximum and minimum

voltage points are sent to the knowledge based to calculate the sensitivity factors.


111

After the control algorithm module determines the required reactive power or active

power curtailment from the DG agent, two messages are sent out via the interface: the

new power set point to the actuator and the action confirmed message to the

information-sharing platform


1) Reactive power control algorithm

The aim of reactive power control by the DG agent is similar to that of the SOP agent.

The objective function is given as

Minimize 𝐽(𝑄𝑔𝑛𝑒𝑤) = ∑ (𝑉𝑖 − 𝑉𝑛𝑜𝑟𝑚)2𝑁

𝑖=1 + 𝜔𝑎𝑏𝑠𝑜𝑟 ∙ 𝑚𝑖𝑛{0, 𝑄𝑔𝑛𝑒𝑤} (5-28)

subject to

|𝑄𝑔𝑛𝑒𝑤| ≤ 𝑄𝑆_𝑖𝑛𝑗

𝑙𝑖𝑚 , 𝑄𝑆_𝑖𝑛𝑗𝑙𝑖𝑚 = √𝑆𝑔

𝑟𝑎𝑡𝑒2− 𝑃𝑔

2 (5-29)



− 𝐾𝑖 ∙ (𝑄𝑔𝑛𝑒𝑤 − 𝑄𝑔

𝑜𝑙𝑑) (5-30)


𝑏𝑎𝑛𝑑 (5-31)

where 𝑄𝑔𝑛𝑒𝑤 is the reactive power supplied by the SOP, and it turns to be a negative

value when the DG absorbs reactive power from the connected feeder; 𝜔𝑎𝑏𝑠𝑜𝑟 is the

weight for absorbing reactive power from the connected feeder; 𝐾𝑖 is the sensitivity

factor for the change of voltage 𝑉𝑖 with respect to the change of reactive power.

2) Active power curtailment algorithm

The objective of this method is to determine the required active power curtailment that

keeps the maximum voltage within the specified upper limits using

𝑉𝑈𝑝𝑝𝑒𝑟2 − 𝑉𝑚𝑎𝑥

𝑜𝑙𝑑 2+ 𝑀𝑚𝑎𝑥 ∙ (𝑃𝑔

𝑛𝑒𝑤 − 𝑃𝑔𝑀𝑃𝑃𝑇) =0 (5-32)


112

where 𝑀𝑚𝑎𝑥 is the sensitivity factor of the maximum voltage point; 𝑃𝑔𝑀𝑃𝑃𝑇 is the

active power generation of the DG units under MPPT mode.

C. Coordination among multiple DG agents

The aim of coordinating multiple DG agents for voltage control is to minimize the cost

of reactive power control and active power curtailment for all participated DG agents.

In this work, all DG agents are assumed to have equal weights of real and reactive

power control for voltage control. Therefore, the cost is only determined by the amount

of reactive power support or active power curtailment of the participated DG agents.

The coordination schemes among DG agents for different control methods are

summarized as follows.

1) Coordination for reactive power control

A prioritization of the DG agents for reactive power control is determined based on the

sensitivity factors using the follows steps:

Step 1) If the information–sharing platform determines to activate the reactive

power control from DG agents, the last downstream DG along the feeder

(most sensitive to voltage change) is informed first to achieve the

optimization purpose. The required reactive power is calculated using

(5-28) and the new reference is updated.

Step 2) If there is still voltage violations after activating the last downstream DG

agent, the next upstream DG agent (second most sensitive to voltage

change) is activated.

Step 3) Step 1) and 2) are repeated until there is no voltage violation along the

feeder or the fist upstream DG agent is reached.

2) Coordination for active power curtailment

For active power control, the DG agent where the connecting bus has an overvoltage is

activated first. The next downstream DG agent is activated if there is still a voltage

violation. This sequence is due to the factors that 1) any DG agent connecting to the


113

downstream feeder of the overvoltage bus has the same while maximum sensitivity

factor to the overvoltage bus; and 2) the further the downstream DG is to the

substation, the more sensitive to voltage change it will be thus has a higher possibility

to cause the voltage at any other bus exceeding the lower limit due to the active power

curtailment.

Voltage control for the OLTC agent

It was assumed in this study that the substation transformer has an on-load tap changer

on the primary winding and the voltage on the primary side remains constant.


The OLTC agent receives the action command and the overall maximum and minimum

voltages of the network from the information-sharing platform. The action command

specifies the control strategy required: feasible solution strategy or infeasible solution

strategy. After the control algorithm module determines the new tap setting of the

OLTC agent, two messages are sent out via the interface: the new reference to the

actuator and the task confirmed message to the information-sharing platform


The control algorithm module receives the voltage data and action command from the

interface and then implements the relevant control strategy described below.

1) Feasible solution strategy

When the agent receives a feasible solution message, i.e., 𝑉𝑚𝑎𝑥 − 𝑉𝑚𝑖𝑛 ≤ 𝑉𝑢𝑝𝑝𝑒𝑟 −

𝑉𝑙𝑜𝑤𝑒𝑟, the control algorithm module calculates the number of required taps at each

time step i using

{𝑇𝑖 = 𝑇𝑖−1 + round (

𝑉𝑚𝑖𝑛−𝑉𝑙𝑜𝑤𝑒𝑟

∆𝑉) , if 𝑉𝑙𝑜𝑤𝑒𝑟 > 𝑉𝑚𝑖𝑛

𝑇𝑖 = 𝑇𝑖−1 + round (𝑉𝑢𝑝𝑝𝑒𝑟 − 𝑉𝑚𝑎𝑥

∆𝑉) , if 𝑉𝑢𝑝𝑝𝑒𝑟 < 𝑉𝑚𝑎𝑥

(5-33)


114

subject to

𝑉𝑠 = 𝑉𝑝 + 𝑇 ∙ ∆𝑉 (5-34)

𝑇upp𝑒𝑟 ≤ 𝑇𝑖 ≤ 𝑇Lower (5-35)

Where 𝑇𝑖 and 𝑇𝑖−1 is the tap position at time step i and i-1; 𝑉𝑚𝑎𝑥 and 𝑉𝑚𝑖𝑛 are the

overall maximum and minimum voltage of the network; 𝑉𝑢𝑝𝑝𝑒𝑟 and 𝑉𝑙𝑜𝑤𝑒𝑟 are the

standard upper and lower voltage limits; 𝑇𝑜𝑙𝑑 is current tap position; ∆𝑉 is the step

voltage of each tap; 𝑉𝑠 is the secondary voltage of the substation; 𝑉𝑝 is primary

voltage of the substation; 𝑇upp𝑒𝑟 and 𝑇Lower are the upper and lower boundary of the

OLTC.

2) Infeasible solution strategy

The infeasible solution strategy is required after the double check process while the

infeasible solution still exists. The aim of this strategy is to keep the minimum voltage,

𝑉𝑚𝑖𝑛 at the specified lower limit. Thus the maximum voltage though still violates the

upper limit but is as low as possible, which in turn minimize the required curtailment of

the DG active power. Thus the new tap setting is obtained using

𝑇𝑖 = 𝑇𝑖−1 + round (𝑉𝑚𝑖𝑛−𝑉𝑙𝑜𝑤𝑒𝑟

∆𝑉) (5-36)

subject to

𝑇upp𝑒𝑟 ≤ 𝑇𝑖 ≤ 𝑇Lower (5-37)

Voltage control for the double check process

The purpose of the double check process is to avoid unnecessary DG active power

curtailment. It is achieved by enabling a feasible solution for the OLTC control by

either controlling the violated extreme voltage or controlling the other extreme voltage.

The whole process is performed within the information-sharing platform as follows:

Step 1) Check for the condition of an infeasible solution before activating the


115

OLTC agent using

𝑉𝑢𝑝𝑝𝑒𝑟 − 𝑉𝑙𝑜𝑤𝑒𝑟 < 𝑉𝑚𝑎𝑥 − 𝑉𝑚𝑖𝑛 (5-38)

Step 2) If the infeasible solution is raised, allocate the feeders where 𝑉𝑚𝑎𝑥 and

𝑉𝑚𝑖𝑛 happen and then initiate the double check process. Otherwise, send

an action command to the OLTC agent to activate the feasible solution

strategy.

Step 3) If the double check process is initiated, activate the control agent (or

method) that is connected to the feeder where 𝑉𝑚𝑎𝑥 or 𝑉𝑚𝑖𝑛 happens.

The control method that has already been activated is not involved in this

process.

Step 4) If there are multiple candidates, repeat step 1) - 3) until there is a feasible

solution or all the candidates have been activated. The sequence of

activating a candidate at each time step is based on the proposed

coordination law shown in Figure 5.13, i.e., SOP active power control,

SOP reactive control and finally DG reactive power control.

5.3.3 Implementation procedures of the information-sharing platform

The overall implementation procedures of the information-sharing platform are

summarized below.

Step 1) Update the voltage information after receiving segment voltage data from

the local control agents.

Step 2) Determine the maximum and minimum voltages in each feeder. Then,

check the condition of voltage violation.

Step 3) If there is no voltage violation, go to step 1. Otherwise, allocate the

feeders at which the voltage exceeds the specified limit and then initiate

the coordinated control on these feeders.

Step 4) Send an action command to activate the relevant agent based on the


116

coordination law described in Section 5.3.2.

Step 5) Each time after receiving the action confirmed message from the agent,

update voltage information by repeating step 1) and 2). If there is no

voltage violation or all available control methods have been activated, go

to step 1). Otherwise, go to step 4).

5.4 Discussion: Benefits of the Proposed Control

Strategy

The main advantages of the coordinated voltage control strategy are summarized

below.

The proposed voltage control framework allows for required coordination

between multiple agents (devices) while independent real-time control by each

local control agent. Moreover, it is easy to modify and upgrade each local

control agent or implement expansion (e.g. more DG connections) without

interrupting other parts of control framework.

Further flexibility is provided by the SOP when the over- and under-voltage

issues exist simultaneously within a network that cannot be solved by the OLTC.

The use of SOP also reduces network power losses compared to reactive power

control (e.g. capacitor banks). Moreover, the power electronic based device

provides fast enough voltage control compared with the tap changing devices

(e.g., OLTC, capacitor banks), which may be needed for DGs to ride through the

voltage sags during emergency conditions.

A trade-off of the objectives, according to their prioritisation, is achieved by

using the proposed priority-based coordination method. This method performed

by an information-sharing platform is in line with the distributed control

concept while unnecessary negotiation process among agents is eliminated and

data communications are reduced.


117

The computational burden is reduced because 1) a distributed voltage control is

achieved by the local control agent, instead of doing a centralised

calculation/optimisation using all data together, and 2) with the use of the

simplified estimation method, the knowledge of the network voltage profile is

obtained by each agent that find their respective maximum voltage and

minimum voltages (within its corresponding segments) simultaneously. In

addition, the numerical convergence problems of running the power flow

algorithm is eliminated since there is no need for running the power flow

algorithm in each time step, especially in the case of high R/X ratio networks.

The communication investments are reduced since the meters and sensors are

placed at the OLTC, DG and SOP connecting points and only the critical data

are exchanged via the information-sharing platform.

5.5 Case Study

The effectiveness of the coordinated voltage control strategy was verified on an MV

distribution network obtained from [55], as shown in Figure 5.16. This network is a

33-bus 11-kV distribution system with four feeders. The four feeders have different

lengths, ratings and load types. Intermittent DG units with different characteristics were

connected to different buses to simulate the unbalance in load/generation diversity

between feeders. Table 5.1 shows a summary of the loads and DGs data. Figure 5.17

depicts the normalized profiles for different load types (commercial, residential and

industrial) and for different DG types within two days (a weekday and a weekend).

These normalized profiles (half-hourly per step) were multiplied by the rated power of

each load or DG (shown in Table 5.1) to have the power daily profile of each bus. All

these data were obtained from [55]. More detailed network data can be found in the

Appendix B. The voltage limits of this test network are 11 kV ±5%. The OLTC was

assumed to have 32 taps with 0.00625 p.u. per step. An SOP was assumed to be

installed between feeder 2 and 3. The capacity limit of the SOP is 1 MVA. This study


118

was programmed in Java and C. Balanced three-phase feeders were considered.

SOP

OLTC

1 2 3 4 5 6 7 8 9 10

1 2 3 4 5 6

1 2 3 4 5 6 7 8 9

1 2 3 4 5 6 7 8

Feeder 1

Feeder 2

Feeder 3

Feeder 4

Wind

PV

33 kV/11 kV

Transformer

0

Figure 5.16: The four-feeder MV test network

Table 5.1: Data of the Load and DG

Feeder

number

Load profile DG profile

Types

Total rated

loading

(MVA)

Types

Locations Ratings

(MVA)

1 Commercial 4 Wind 3, 8 2, 0.5

2 Residential 5 PV 2, 5, 10 5, 2, 1

3 Industrial 3 Wind 3, 6 0.5, 2.5

4 Residential 4.5 PV, wind 5, 8 0.5, 0.5

(a) (b)

Figure 5.17: Normalized profiles of a weekend and weekday: (a) loads; (b) DGs

0

0.2

0.4

0.6

0.8

1

No

mal

ized

act

ive

po

wer

(P

g)

Time (hours)

Wind PV

6am 6pm6am 6pm12pm 12pm12am 12am 12am0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

No

min

aliz

ed p

ow

er

Time (hours)

CommertialResidentialIndustrial

6am 6pm6am 6pm12pm 12pm12am 12am 12am


119

5.5.1 Voltage Profile Estimation

The voltage profile generated by the estimation method proposed in section 5.2.1 was

compared with the actual voltage profile (calculated by power flow analysis). Figure

5.18 shows the overall maximum and minimum voltages along each feeder during the

studied two days before voltage control. The maximum voltage was read from meters

thus the same. It can be seen that the estimated minimum voltages were close to the

actual values (the maximum difference is 0.004 p.u). For the minimum voltages, lower

voltage bounds were generated by the proposed simplified estimation method (see the

red curves). Therefore, more conservative voltage references will be used for voltage

control using the proposed estimation method.

(a) (b)

0.87

0.895

0.92

0.945

0.97

0.995

1.02

1.045

Vm

axan

d V

min

of

feed

er 2

(p

.u.)

Time (Hour)

V_estimate-Feeder 2

V_actual-Feeder 2

12am 6am 6pm12pm 12am 6am 6pm12pm 12am0.92

0.935

0.95

0.965

0.98

0.995

1.01

1.025

Vm

axan

d V

min

of

feed

er 1

(p

.u.)

Time (Hour)

V_estimate-Feeder 1

V_actual-Feeder 1

12am 6am 6pm12pm 12am 6am 6pm12pm 12am


120

(c) (d)

Figure 5.19 shows the snapshot voltage profiles of all buses along each of the four

feeders at 11 am in the first studied day. The voltage profile by the proposed method

were generated by connecting the critical voltage points, i.e., the measured voltages (at

the substation busbar, SOP and DG connecting buses) and the estimated minimum

voltage within each segment using the segment estimation method. It is clear that all

the maximum and minimum voltage points were captured as expected.

(a) (b)

0.955

0.965

0.975

0.985

0.995

1.005

1.015

1.025

1.035

Vm

axan

d V

min

of

feed

er 3

(p

.u.)

Time (Hour)

V_estimated-Feeder 3V_actual-Feeder 3

12am 6am 6pm12pm 12am 6am 6pm12pm 12am0.9

0.91

0.92

0.93

0.94

0.95

0.96

0.97

0.98

0.99

1

Vm

axan

d V

min

of

feed

er 4

(p

.u.)

Time (Hour)

V_estimate-Feeder 4V_actual-Feeder 4

12am 6am 6pm12pm 12am 6am 6pm12pm 12am

Figure 5.18: Per unit maximum and minimum voltage profile over the studied period

at (a) feeder 1; (b) feeder 2; (c) feeder 3; (d) feeder 4

0.985

0.99

0.995

1

1.005

1.01

1.015

1.02

0 1 2 3 4 5 6 7 8

Vo

ltag

e (p

.u.)

Buses of feeder 1

Voltage profile by power flow analysis

Volatge profile by proposed method

0.98

0.985

0.99

0.995

1

1.005

1.01

1.015

1.02

0 1 2 3 4 5 6 7 8 9 10

Vo

ltag

e (p

.u.)

Buses of feeder 2




121

(c) (d)

5.5.2 Coordinated voltage control using OLTC and SOP

Most utilities prefer DG units to operate close to the unity power factor and do not

allow regulating the voltage actively at the Point of Common Coupling [103]. In this

context, the coordinated control merely considering the OLTC and SOP was conducted

to investigate the performance of the proposed control strategy. All the DG units were

assumed to operate in maximum power point tracking (MPPT) with unity power factor.

Four control schemes were carried out for comparisons:

Scheme I: Voltage control using the OLTC;

Scheme II: Coordinated control between the OLTC and the SOP assuming only

active power control;

Scheme III: Coordinated control between the OLTC and the SOP assuming only

reactive power control;

Scheme IV: Coordinated control between the OLTC and the SOP assuming both real

and reactive power control based on the proposed priority-based

coordination.

0.94

0.95

0.96

0.97

0.98

0.99

1

1.01

1.02

0 1 2 3 4 5 6 7 8 9

Vo

ltag

e (p

.u.)

Buses of feeder 4



0.98

0.985

0.99

0.995

1

1.005

1.01

1.015

1.02

0 1 2 3 4 5 6

Vo

ltag

e (p

.u.)

Buses of feeder 3



Figure 5.19: Per unit voltage profile at 11am in the first studied day along (a) feeder 1;

(2) feeder 2; (3) feeder 3; (4) feeder 4


122

The performance of the control schemes were evaluated in terms of the network voltage

profile, the number of tap operations, total network power losses and the required size

for the SOP to achieve voltage control.

Network voltage profile

Figure 5.20 shows the overall maximum and minimum voltages of the network during

the studied period when different control schemes are used. As shown in the figure, the

network voltage profiles were within the specified limits for most of the time by using

the OLTC, i.e., Scheme I. However, it could not solve the areas of infeasible solutions

shown in the figure. These areas of infeasible solutions were mitigated when the SOP

was coordinated with the OLTC, i.e., using Scheme II, III and IV.

Figure 5.20: maximum and minimum voltage profiles at different control schemes

Tap operation

Figure 5.21 compares the number of tap changes during the studies period when

different control schemes are used. As shown in the figure, the number of tap

operations was reduced significantly when the SOP was coordinated with the OLTC,

i.e., using Scheme II, III and IV. The highest reduction was achieved by using Scheme

0.9

0.95

1

1.05

1.1

Vm

axan

d V

min

of

the

net

wo

rk (

p.u

.)

Time (Hours)

Scheme I Scheme II Scheme III Scheme IV

12am 6am 6pm12pm 6pm 12am12pm12am 6am

Infeasible solution Infeasible solution Infeasible solution

upper limit

Lower limit


123

IV, which was reduced from 46 to 16 taps compared with Scheme I.

Figure 5.21: Change of the tap operation at different control schemes

Network power losses

Figure 5.22 compares the total network losses at different control schemes. It can be

seen that the total network losses were reduced when Scheme II and IV were used. The

highest loss reduction (by 21.93%) was achieved when the active power control of the

SOP was merely considered (Scheme II). Such loss reduction benefit was slightly

lessened by 19.41% when Scheme IV was used due to the participation of reactive

power control. However, on the contrary, the total network power losses were increased

by 11.60% when the reactive power control of the SOP was merely considered (Scheme

III). This is due to the excessive absorption of reactive power from the network, which

is in line with the previous results obtained in Section 5.3.1.

-1

1

3

5

7

9

11

13

Tap

po

siti

on

Time (hours)




124

Figure 5.22: Total network losses at different control schemes

Required size of the SOP

Figure 5.23 shows the required size (in apparent power) of the SOP for the

corresponding control schemes. All the three control schemes considering SOP

compensate the changes of the load demands and DG productions which, in turn,

reduces the dynamic changes of the voltage profiles and hence the number of tap

operations. However, the SOP operated at its rated power for most of the time when

only controls reactive power (Scheme III). When controls active power (Scheme II), the

SOP operated at its rated power for only a few hours. The total hours, operating at rated

power using the coordinated real and reactive power control (Scheme IV), was in

between the other two.

50

150

250

350

450

550

650

750

850

Tota

l net

wo

rk lo

sses

(kW

)

Time (Hours)




125

Figure 5.23: Apparent power requirements for different SOP control schemes

From the above results, it is concluded that the coordination between the OLTC and the

SOP is essential when the OLTC is infeasible to solve the voltage problems. In addition,

the proposed priority-based coordination of the OLTC and SOP achieves the best

compromise of the sub-objectives (in accordance with the prioritisation of the

objectives). It enables smaller number of tap operations comparing to the scheme of

only using the active power control and fewer power losses comparing with only using

the reactive power control.

5.5.3 Coordinated voltage control using OLTC, SOP and DG

Increasing the DG penetration in the distribution network brings more difficulties in

controlling network voltages. The previous control using the OLTC and the SOP may

not be able to control the voltage within the specified limits in such a case. Figure 5.24

shows the overall maximum and minimum voltages of the network during the studied

period when the capacity (rated power shown in Table 5.1) of the DG units (connected

in feeder 2) were enlarged by 10%. It is shown that an overvoltage problem still exist

during the peak PV generation when the coordinated control between OLTC and SOP

was used.

0.25

0.5

0.75

1

S so

p (M

VA

)

Time (Hours)

Scheme II Scheme III Scheme IV



126

Figure 5.24: maximum and minimum voltage profiles when increase DG penetration

The coordinated control using the OLTC, SOP and DG units was conducted in this part.

It was assumed that DG units installed at bus 9 in feeder 1, bus 5 in feeder 2, bus 6 in

feeder 3 and bus 8 in feeder 4 were inverter-based DG unit and thus had the reactive

power control capability. All DG owners were obligated to respond to the active power

curtailment request.

Two different cases were carried out for comparison purposes, which are the

coordinated control using OLTC and DG units and the coordinated control using OLTC,

SOP and DG units. The performance of these two cases were evaluated in terms of the

network voltage profile, the total active power generated by DGs, the number of tap

operations, total network power losses and the power settings of the SOP and DG unit.

Network voltage profile and total active power generated by DGs

Figure 5.25 shows the maximum and minimum voltages of the network during the

studied period. It can be seen that the overvoltage problem was solved for both cases

when the DG units were involved in voltage control. However, DG active power

curtailment was required for the case without involving the SOP control, as shown in

0.9

0.95

1

1.05

1.1

Vm

axan

d V

min

of

the

net

wo

rk (

p.u

.)

Time (Hours)

Coordinated control using OLTC and SOP


Overvoltage

Upper limit

Lower limit


127

Figure 5.26. There was no curtailed DG active power when OLTC, SOP and DG units

were coordinated for voltage control.

Figure 5.25: maximum and minimum voltage profiles with increased DG penetration

by coordinated control with and without involving SOP

Figure 5.26: Total active power generated with increased DG penetration by

coordinated control with and without involving SOP

Tap operation

Figure 5.27 compares the number of tap changes during the studied period at different

cases. As shown in the figure, a smaller number of tap operation was required when the

OLTC, SOP and DG units were coordinated. The tap operation was reduced from 21 to

0.9

0.95

1

1.05

1.1

Vm

axan

d V

min

of

the

net

wo

rk (

p.u

.)

Time (Hours)

Coordinated control using OLTC and DGs Coordinated control unsing OLTC, SOP and DGs


Upper limit

Lower limit

Curtailed energy

Curtailed energy

To

tal

acti

ve

po

wer

(k

W)


128

15 taps.

Figure 5.27: Change of the tap operation with increased DG penetration at different

cases.

Network power losses

Figure 5.28 compares the total network losses during the studied period at difference

cases. It can be seen that the total network losses were reduced by 18.98% when the

SOP were involved in the coordinated voltage control.

Figure 5.28: Total network losses with increased DG penetration

-2

0

2

4

6

8Ta

p p

osi

tio

n

Time(hours)

Coordinated control using OLTC and DGs Coordinated control considering OLTC, SOP and DGs


50

150

250

350

450

550

650

750

850

Tota

l net

wo

rk lo

sses

(kW

)

Time(Hours)

Coordinated control using OLTC and DGs

Coordinated control using OLTC,SOP and DGs



129

Power settings of the SOP and DG unit

Figure 5.29 shows the required size (in apparent power) for the SOP and the reactive

power for the PV DG units at bus 5 in feeder 2 during the studied period. As shown in

Figure 5.29a, the SOP operates at its rated power for most of the time. For the PV DG

unit, as shown in Figure 5.29b, it was operated under its rated power for most of the

time when the SOP was not involved in the voltage control. However, the amount of

reactive power support from the DG unit was reduced significantly when OLTC, SOP

and DG were coordinated for voltage control.

(a)

(b)

Figure 5.29: Power settings with increased DG penetration: (a) apparent power

settings of the SOP, (b) reactive power settings of the PV DG unit

0.25

0.5

0.75

1

S SO

P (M

VA

)

Time (Hours)12am 6am 6pm12pm 6pm 12am12pm12am 6am

0

0.5

1

1.5

2

2.5

Qg

(MV

AR

)

Time (Hours)

Coordinated control using OLTC and DGs

Coordinated control using OLTC, SOP and DGs



130

From the above results, it is concluded that the proposed coordination using the OLTC

DG units and SOP is desirable in solving the voltage violation problem under high DG

penetration while satisfying the sub-objectives, i.e., less active power curtailment,

reduced number of tap operations and network power losses. The use of the proposed

SOP control also reduces the amount of reactive power support from DG units for

voltage control.

5.6 Summary

A novel coordinated voltage control strategy was proposed to maintain the network

voltages within the specified limits whilst mitigating DG active power curtailment,

reducing tap operations and network power losses. The OLTC, the SOP and multiple

DG units were considered as the voltage control devices in this study.

A coordinated but distributed voltage control framework was developed which uses a

simplified voltage profile estimation method. The simplified voltage profile estimation

method obtain the maximum and minimum voltages of each feeder based on the

measurements at the substation substation busbar, the DG and SOP locations and thus

without having to measure the voltage at each network node. The voltage control

framework is based on a distributed coordination where the OLTC, the SOP and the DG

units are considered as the local control agents. These agents operate independently

based on the local measurements and achieve the coordination via data exchange

through an information-sharing platform.

Followed by an elaborated analysis of using SOP for distribution network voltage

control, a priority-based coordination using the proposed voltage control framework

was developed. This coordination allows the local control agents to be activated in a

predefined sequence to achieve a trade-off among the objectives according to their

prioritisation while remaining the independence of each agent.


131

Simulations verified the effectiveness of the proposed coordinated control strategy

under various load and generation conditions as well as high DG penetration. The

results showed that the proposed coordinated strategy has the capability to control

network voltage while also mitigating DG active power curtailment, reduce the tap

operations and avoid unnecessary increase in power losses in accordance with their

prioritisation. Simulation results also illuminated the role and importance of the SOP

for distribution network voltage control. It was found that the SOP is capable of

compensating the OLTC control, avoiding unnecessary DG curtailment, reducing the

number of tap operations and the network power losses as well as reducing the amount

of reactive power support from DG units for voltage control.

Chapter 6 Conclusions and Future Work

132

Conclusions and Future Work

HIS chapter outlines the major contributions and findings from the

research. It also presents further work around the research areas in this

thesis.

Chapter 6

T


133

6.1 Conclusions

Soft Open Points are power electronic devices installed in place of normally-open

points in electrical power distribution networks. They are able to provide active power

flow control, reactive power compensation and voltage regulation under normal

network operating conditions, as well as fast fault isolation and post-fault supply

restoration under abnormal conditions. The use of SOPs for the operation of MV

distribution networks was investigated from different perspectives including the control

of a back-to-back VSC based SOP, the benefits analysis of using SOPs and distribution

network voltage control with SOPs.

6.1.1 Control of a Back-to-Back VSC based SOP

Two control modes were developed for the operation of an SOP based on back-to-back

VSCs. The power flow control mode regulates both active and reactive power flows on

the connected feeders using a dual closed-loop current-controlled strategy. The supply

restoration mode provides fast post-fault supply restoration to the isolated loads using a

voltage and frequency control strategy. The operational principle and performance of

the back-to-back VSC based SOP with these two control modes were analysed under

both normal and abnormal network operating conditions:

Under normal conditions, the proposed SOP controller is able to provide both

instantaneous and longer duration power flow regulation between the connected

feeders as well as dynamic Volt/VAR control on both interface terminals.

During fault conditions, the SOP controller (power flow control mode) is able to

isolate both balanced and unbalanced network fault between the connected

feeders. Thus the fault propagation on the network can be limited and the

short-circuit current can be maintained as same as the radial network operation

without the SOP, which is important from a practical point of view as there is

little impact on the existing protection schemes.


134

Under post-fault supply restoration conditions, fast supply restoration (cold load

pickup) can be achieved through a smooth transition between the power flow

control mode and the supply restoration mode. For the transition from the power

flow control mode to the supply restoration mode, a soft cold load pickup process

was found preferred than a direct mode switching in order to avoid DC offset and

inrush issues of current which may trigger unexpected protection operations. For

the transition from the supply restoration mode to the power flow control mode,

the undesirable current impulse and voltage drop can be reduced after using the

voltage synchronization procedure and resetting the PI controllers.

6.1.2 Benefit Analysis of SOPs for MV Distribution Network

Operation

The operational benefits of using SOPs in MV distribution networks were investigated

focusing on power loss reduction, feeder load balancing and voltage profile

improvement. A steady state analysis framework was developed to quantify the benefits

of using SOPs. A generic power injection model of an SOP that is suitable for steady

state analysis was developed. Both physical limits and internal power losses of a typical

SOP device: back to back VSCs were considered in the model. The optimal SOP

operation was obtained using improved Powell’s Direction Set method. A combined

method considering both SOP and network reconfiguration were also proposed to

identify the benefits.

Based on the proposed framework, quantitative and sensitive analysis were conducted

and the key findings are listed as follows:

SOPs can contribute to significant power loss reduction, feeder load balancing

and voltage profile improvement. In the case study, using only one SOP achieved

similar improvement on network power loss reduction and feeder load balancing

compared to the case of using network reconfiguration with all branches


135

equipped with remotely controlled switches. By combining SOP and network

reconfiguration, the greatest improvements were obtained and smaller SOP sizes

were required.

When a large capacity of DGs were connected to the distribution network, SOPs

were found be able to significantly reduce the peak currents in feeders and

alleviate undesirable voltage excursions induced by the connection of DG and

demand. Therefore, SOPs can be used as an alternative to infrastructure upgrades

in accommodating distributed energy resources.

When the device efficiency (at rated power) decreased, the economic benefit of

SOP obtained from reducing total network power losses were lowered. However,

a greater positive impact were obtained when the system loading increased. The

requirements on SOP device efficiency for network loss reduction were reduced

when the system loading increased.

6.1.3 Voltage Control in Active Distribution Networks with SOPs

A novel coordinated voltage control strategy was proposed, in which OLTC, SOP and

DG units were considered as the controllable devices. The objective of the control

strategy was to maintain the network voltage within the specified limits whilst avoiding

unnecessary DG active power curtailment, reducing tap operations and network power

losses.

A coordinated but distributed voltage control framework was developed which uses a

simplified voltage profile estimation method. The simplified voltage profile estimation

method obtains the maximum and minimum voltages of each feeder based on the

measurements at the substation busbar, the DG and SOP locations and thus without

having to measure the voltage at each network node, which is able to reduce the

investment cost. The voltage control framework is based on a distributed control

scheme where the OLTC, the SOP and the DG units are considered as the local control


136

agents. These agents operate independently based on the local measurements and

achieve the coordination via data exchange through an information-sharing platform.

Followed by an analysis of using SOP for distribution network voltage control, a

priority-based coordination using the proposed control framework was developed. This

coordination, without affecting the independence of each agent, allows the local control

agents to be activated in a predefined sequence that can achieve a trade-off among the

objectives according to their prioritisation.

The results presented prove the capability of the proposed coordinated strategy to

achieve a trade-off among objectives in accordance with their prioritisation. Simulation

results also illuminated the role and importance of the SOP for distribution network

voltage control. It was found that the SOP is capable of compensating the OLTC

control, avoiding unnecessary DG active power curtailment, reducing the number of tap

operations and the network power losses as well as reducing the amount of reactive

power support from DG units for voltage control.

6.2 Future Work

The following future work was identified to extend the work reported in this thesis:

6.2.1 Use of other control strategies for the operation of the

back-to-back VSC based SOP

In this study, when a fault occurs on a feeder, a voltage and frequency control strategy

was used for the SOP to support isolated loads due to fault isolation. However, this

control strategy cannot limit the overcurrent on the interfacing VSC when a further

fault occurs on the isolated area. Hence additional protection of the interfacing VSC

needs to be considered. Therefore, control strategies that can provide the function of

limiting overcurrent should be investigated. One possible solution is to add an inner


137

current control loop to the current control strategy to limit the fault current.

A mode switching method were used to achieve a transition between the two control

modes of the SOP. The parameters of the PI controllers need to be determined carefully

upon the switching of the control modes in order to avoid unexpected transient

oscillations on the output voltage and current. Control strategies of the SOP that can

avoid the mode switching process should be investigated to improve the reliability.

6.2.2 Performances of the back-to-back VSC based SOP considering

different load types

The performances of a back-to-back VSC based SOP in MV distribution networks have

been investigated considering passive loads with constant impedances. However,

induction and/or synchronous machines (e.g., induction motor, synchronous generator)

are usually connected to distribution networks. These types of loads have different load

characteristics and, hence, may lead to different transient behaviours. Therefore, further

investigations on the performance of the proposed SOP device with different types of

network loads should be undertaken as future work.

6.2.3 Impacts of SOPs on feeder automation

Feeder automation is realized through automatic detection, isolation and

reconfiguration of faulted segments on distribution feeders. In this study, the capability

of the back-to-back VSC based SOP for fault isolation and post-fault supply restoration

was investigated. However, the interactions between the SOP and existing feeder

automation schemes were not considered.

For example, a simple form of feeder automation is achieved by deploying

combinations of reclosers and sectionalizers in a radial feeder. NOPs are used to allow

the radial feeders operating as an ‘open ring’. These devices (reclosers, sectionalizers

and NOPs) are coordinated to minimize the interruption of supply to customers under


138

fault conditions. When the NOPs are replaced with SOPs, the impact of the SOP on the

existing feeder automation scheme should be assessed. Additionally, the coordination

between the SOP and existing feeder automation scheme to avoid incoherent control

actions and approaches to obtain an optimal solution should be investigated.

6.2.4 Use of SOPs in unbalanced three-phase distribution networks

In this thesis, the use of SOPs in balanced three-phase distribution networks was

investigated. However, the domestic and some commercial customers connected to the

distribution network usually have single-phase connections. Therefore, the load in

distribution networks cannot be perfectly balanced. The imbalance in the three-phase

equipment needs to be reduced.

The SOP contains a common DC bus and hence enabling exchange of instantaneous

power between phases, which can be used to rebalance the power flows on the

connected feeders. Hence, new controllers should be investigated for the operation of

the SOP to address unbalanced three-phase conditions. In addition, the benefits of using

SOPs for the operation of unbalanced three-phase distribution networks should be

quantified and evaluated. Single phase power flow and optimization are suggested as

possible solution methods for the benefit analysis. New voltage control strategy to

address the voltage imbalance among the three phases should also be investigated.

6.2.5 Economic analysis of SOPs in MV distribution networks

This thesis investigated the technical benefits of SOPs in MV distribution networks.

However, a purely economic evaluation will give a better idea of weather the SOP

implementation is feasible, and where costs must be reduced to make it feasible. Hence,

further study on the economic analysis of SOPs should be undertake as future work.

The life cycle costs of SOPs should be quantified. The optimal amount, locations and

sizes of SOPs should also be determined considering the capital cost of SOPs. In


139

addition, the power loss reduction benefit of SOPs was performed in this thesis. In

order to further quantify the loss reduction benefit of SOPs, this study should be

extended to the study of quantifying energy power losses during a period.

Reference

140

Reference

[1] National Aeronautics and Space Administration, "The Current and Future

Consequences of Global Change", 08-Aug-2015. [Online]. Available:

http://climate.nasa.gov/effects/. [Accessed: 11-Sep-2015]

[2] T. Overbye et al., "The Electric Power Industry and Climate Change: Power Systems

Research Possibilities", Jun-2007. [Online]. Available:

http://pserc.wisc.edu/documents/publications/reports/2007_reports/pserc_climate_chan

ge_final_rpt_june07.pdf. [Accessed: 11-Sep-2015]

[3] European Commission, "EU action on climate", 15-July-2015. [Online]. Available:

http://ec.europa.eu/clima/policies/brief/eu/index_en.htm. [Accessed: 11-Sep-2015]

[4] Department of Energy and Climate Change, UK, "2010 to 2015 government policy:

greenhouse gas emissions", 08-May-2015. [Online]. Available:

https://www.gov.uk/government/publications/2010-to-2015-government-policy-greenh

ouse-gas-emissions/2010-to-2015-government-policy-greenhouse-gas-emissions.

[Accessed: 11-Sep-2015]

[5] Her Majesty's Goverment, UK, "The UK Renewable Energy Strategy", Jul-2009.

[Online]. Available:

http://www.biomassenergycentre.org.uk/pls/portal/docs/PAGE/RESOURCES/REF_LI

B_RES/PUBLICATIONS/THEUKRENEWABLEENERGYSTRATEGY2009.PDF.


[6] K. Qian, "Modelling of electrical distribution systems with high penetration level of

distributed generation and electric vehicles," Ph.D thesis, Glasgow Caledonian

University, 2000.

[7] C. L. Masters et al., "Statistical evaluation of voltages in distribution systems with

embedded wind generation," IEE Proc. Gener. Transm. Distrib., vol. 147, no.4, pp.

207-212, Jul. 2000.

[8] I. P. D. Povh, D. Retzmann, and M. G. Weinhold, "Future Developments in Power

Industry," presented at the 4th IERE General Meeting and IERE Cental and Eastern

Europe Forum, Poland, 2014.

[9] International Energy Agency, "World Energy Outlook" 2009 [Online]. Available:

www.worldenergyoutlook.org/media/weowebsite/2009/WEO2009.pdf. [Accessed:

11-Sep-2015]

[10] Electricity North West, UK, "Capacity and carbon challenges", 2014. [Online].

Available:

http://www.enwl.co.uk/c2c/future-challenges/capacity-and-carbon-challenges

[11] Electricity North West, UK, "Impact of low carbon future", 2014. [Online]. Available:

http://www.enwl.co.uk/c2c/future-challenges/impact-of-a-low-carbon-future.


[12] Ofgem, "Project Discovery Energy Market Scenarios", consultation document,

09-Oct-2009. [Online]. Available:

Reference

141

https://www.ofgem.gov.uk/ofgem-publications/40361/discoveryscenarioscondocfinal.p

df. [Accessed: 11-Sep-2015]

[13] H. Farhangi, "The path of the smart grid," Power and Energy Magazine, IEEE, vol. 8,

pp. 18-28, 2010.

[14] J. Zhu, Power System Operation Optimization, Second ed., Wiley-IEEE Press, 2008.

[15] U.S. Department of Energy, "Smart Grid System Report", Jul- 2009, [Online].

Available:

http://www.oe.energy.gov/sites/prod/files/oeprod/DocumentsandMedia/SGSRMain_09

0707_lowres.pdf. [Accessed: 11-Sep-2015]

[16] Department of Energy and Climate Change, UK, "Smart Grid Vision and Routemap

Smart", 27-Feb-2014. [Online]. Available:

https://www.gov.uk/.../Smart_Grid_Vision_and_RoutemapFINAL.pdf. [Accessed:

11-Sep-2015]

[17] Electricity Networks Strategy Group, UK, "'A Smart Grid Routemap’ Executive

Summary", Feb-2010. [Online]. Available:

http://webarchive.nationalarchives.gov.uk/20100919181607/http:/www.ensg.gov.uk/ass

ets/smartgrid_routemap_executive_summary_final.pdf. [Accessed: 11-Sep-2015]

[18] T. G. Richard Silversides, Tom Luth. "Power Electronics in Distribution System

Management", HubNet Position Paper, 05-Jan-2015. [Online]. Available:

www.hubnet.org.uk/filebyid/633/PE_Distribution.pdf. [Accessed: 11-Sep-2015]

[19] J. Ekanayake, N. Jenkins, K. Liyanage, J. Wu, and A. Yokoyama, Smart Grid:

Technology and Applications, Wiley Press, 2012.

[20] B. A. Robbins, C. N. Hadjicostis, and A. D. Dominguez-Garcia, "A Two-Stage

Distributed Architecture for Voltage Control in Power Distribution Systems," Power

Systems, IEEE Transactions on, vol. 28, pp. 1470-1482, 2013.

[21] European Network of Transmission System Operators, "Network Code on Operational

Security", 24-Sep-2013 [Online]. Available:

http://www.acer.europa.eu/Official_documents/Acts_of_the_Agency/Annexes%20to%

20ACER%20Recommendation%20092013%20on%20the%20NC%20on/The%20ame

nded%20Network%20Code%20on%20Operational%20Security%20resubmitted%20o

n%2024%20September.pdf. [Accessed: 11-Oct-2015].

[22] G. Strbac and J. Mutale, "Framework and Methodology for Pricing of Distribution

Networks with Distributed Generation - A report to OFGEM", Mar-2005 [Online].

Available:

https://www.ofgem.gov.uk/ofgem-publications/44458/10147-strbacmutale.pdf.

[Accessed: 11-Oct-2015].

[23] N. Hodge, "Power trip", 2010. [Online]. Available:

http://www.agcs.allianz.com/assets/PDFs/GRD/GRD%20individual%20articles/Power

_blackout_risks_article.pdf. [Accessed: 11-Oct-2015].

[24] T. Green et al., "Issues for Distribution System Operation at 2030", HubNet Position

Paper, 11-May-2014. [Online]. Available:

http://www.hubnet.org.uk/filebyid/523/2030GridOperation.pdf. [Accessed:

11-Oct-2015].

http://www.acer.europa.eu/Official_documents/Acts_of_the_Agency/Annexes%20to%20ACER%20Recommendation%20092013%20on%20the%20NC%20on/The%20amended%20Network%20Code%20on%20Operational%20Security%20resubmitted%20on%2024%20September.pdf




http://www.ofgem.gov.uk/ofgem-publications/44458/10147-strbacmutale.pdf

http://www.agcs.allianz.com/assets/PDFs/GRD/GRD%20individual%20articles/Power_blackout_risks_article.pdf

http://www.agcs.allianz.com/assets/PDFs/GRD/GRD%20individual%20articles/Power_blackout_risks_article.pdf

Reference

142

[25] OFGEM, UK, "Losses Incentive Mechanism", 2015. [Online]. Available:

https://www.ofgem.gov.uk/electricity/distribution-networks/losses-incentive-mechanis

m. [Accessed: 11-Oct-2015].

[26] OFGEM, UK, "Decision on restatement of 2009-10 data and closing out the DPCR4

losses incentive mechanism", 21-Mar-2014. [Online]. Available:

https://www.ofgem.gov.uk/ofgem-publications/86757/decisiononclosingoutdprc4losses

mechanism-mar-14.pdf. [Accessed: 11-Oct-2015].

[27] S. Civanlar, J. J. Grainger, H. Yin, and S. S. H. Lee, "Distribution feeder

reconfiguration for loss reduction," Power Delivery, IEEE Transactions on, vol. 3, pp.

1217-1223, 1988.

[28] I. Roytelman and V. Ganesan, "Coordinated local and centralized control in distribution

management systems," Power Delivery, IEEE Transactions on, vol. 15, pp. 718-724,

2000.

[29] R. S. Rao, K. Ravindra, K. Satish, and S. V. L. Narasimham, "Power Loss

Minimization in Distribution System Using Network Reconfiguration in the Presence

of Distributed Generation," Power Systems, IEEE Transactions on, vol. 28, pp.

317-325, 2013.

[30] A. Merlin and H. Back, "Search for a minimal-loss operating spanning tree

configuration in an urban power distribution system," presented at the Proc. 5th Power

System Computation Conforence, Cambridge, 1975.

[31] K. Nara, A. Shiose, M. Kitagawa, and T. Ishihara, "Implementation of genetic

algorithm for distribution systems loss minimum re-configuration," Power Systems,

IEEE Transactions on, vol. 7, pp. 1044-1051, 1992.

[32] D. Das, "A fuzzy multiobjective approach for network reconfiguration of distribution

systems," Power Delivery, IEEE Transactions on, vol. 21, pp. 202-209, 2006.

[33] J. Young-Jae, K. Jae-Chul, O. K. Jin, S. Joong-Rin, and K. Y. Lee, "An efficient

simulated annealing algorithm for network reconfiguration in large-scale distribution

systems," Power Delivery, IEEE Transactions on, vol. 17, pp. 1070-1078, 2002.

[34] Y. Yu and J. Wu, "Loads combination method based core schema genetic shortest-path

algorithm for distribution network reconfiguration," in Proceedings of Power System

Technology Conference, 2002, pp. 1729-1733, vol 3.

[35] A. Ahuja, S. Das, and A. Pahwa, "An AIS-ACO Hybrid Approach for Multi-Objective

Distribution System Reconfiguration," Power Systems, IEEE Transactions on, vol. 22,

pp. 1101-1111, 2007.

[36] C. Chao-Shun, T. Cheng-Ta, L. Chia-Hung, H. Wei-Lin, and K. Te-Tien, "Loading

Balance of Distribution Feeders With Loop Power Controllers Considering

Photovoltaic Generation," Power Systems, IEEE Transactions on, vol. 26, pp.

1762-1768, 2011.

[37] W. Cao, J. Wu, and N. Jenkins. "Feeder Load Balancing in MV Distribution Networks

Using Soft Normally-Open Points,"in IEEE Innovative Smart Grid Technologies

Conference Europe, 2014, PP 1-6.

[38] S. K. Bhattacharya and S. K. Goswami, "Distribution Network Reconfiguration

Considering Protection Coordination Constraints," Electric Power Components and

Reference

143

Systems, vol. 36, pp. 1150-1165, 2008.

[39] S. Rahimi, M. Marinelli, and F. Silvestro, "Evaluation of requirements for Volt/Var

control and optimization function in distribution management systems," in IEEE

International Energy Conference and Exhibition (ENERGYCON), 2012, pp. 331-336.

[40] F. A. Viawan and D. Karlsson, "Voltage and Reactive Power Control in Systems With

Synchronous Machine-Based Distributed Generation," Power Delivery, IEEE

Transactions on, vol. 23, pp. 1079-1087, 2008.

[41] H. E. Z. Farag, E. F. El-Saadany, and R. Seethapathy, "A Two Ways

Communication-Based Distributed Control for Voltage Regulation in Smart

Distribution Feeders," Smart Grid, IEEE Transactions on, vol. 3, pp. 271-281, 2012.

[42] L. Xiaohu, A. Aichhorn, L. Liming, and L. Hui, "Coordinated Control of Distributed

Energy Storage System With Tap Changer Transformers for Voltage Rise Mitigation

Under High Photovoltaic Penetration," Smart Grid, IEEE Transactions on, vol. 3, pp.

897-906, 2012.

[43] M. J. E. Alam, K. M. Muttaqi, and D. Sutanto, "A Multi-Mode Control Strategy for

VAr Support by Solar PV Inverters in Distribution Networks," Power Systems, IEEE


[44] C. Joon-Ho and K. Jae-Chul, "Advanced voltage regulation method of power

distribution systems interconnected with dispersed storage and generation systems,"

Power Delivery, IEEE Transactions on, vol. 16, pp. 329-334, 2001.

[45] F. Shahnia, M. T. Wishart, and A. Ghosh, "Voltage regulation, power balancing and

battery storage discharge control by smart demand side management and

multi-objective decision making," in Power Engineering Conference (AUPEC), 2013

Australasian Universities, 2013, pp. 1-5.

[46] N. Yorino, Y. Zoka, M. Watanabe, and T. Kurushima, "An Optimal Autonomous

Decentralized Control Method for Voltage Control Devices by Using a Multi-Agent

System," Power Systems, IEEE Transactions on, vol. 30, pp. 2225-2233, 2015.

[47] P. M. S. Carvalho, P. F. Correia, and L. A. F. Ferreira, "Distributed Reactive Power

Generation Control for Voltage Rise Mitigation in Distribution Networks," Power


[48] A. Kulmala, S. Repo, Ja, x, and P. rventausta, "Coordinated Voltage Control in

Distribution Networks Including Several Distributed Energy Resources," Smart Grid,


[49] P. N. Vovos, A. E. Kiprakis, A. R. Wallace, and G. P. Harrison, "Centralized and

Distributed Voltage Control: Impact on Distributed Generation Penetration," Power


[50] M. R. Kleinberg, K. Miu, N. Segal, H. Lehmann, and T. R. Figura, "A Partitioning

Method for Distributed Capacitor Control of Electric Power Distribution Systems,"

Power Systems, IEEE Transactions on, vol. 29, pp. 637-644, 2014.

[51] M. B. Liu, C. A. Canizares, and W. Huang, "Reactive Power and Voltage Control in

Distribution Systems With Limited Switching Operations," Power Systems, IEEE


[52] M. A. Azzouz, M. F. Shaaban, and E. F. El-Saadany, "Real-Time Optimal Voltage

Reference

144

Regulation for Distribution Networks Incorporating High Penetration of PEVs," Power


[53] T. Senjyu, Y. Miyazato, A. Yona, N. Urasaki, and T. Funabashi, "Optimal Distribution

Voltage Control and Coordination With Distributed Generation," Power Delivery, IEEE


[54] K. Young-Jin, A. Seon-Ju, H. Pyeong-Ik, P. Gi-Chan, and M. Seung-Il, "Coordinated

Control of a DG and Voltage Control Devices Using a Dynamic Programming

Algorithm," Power Systems, IEEE Transactions on, vol. 28, pp. 42-51, 2013.

[55] H. E. Z. Farag and E. F. El-Saadany, "A Novel Cooperative Protocol for Distributed

Voltage Control in Active Distribution Systems," Power Systems, IEEE Transactions

on, vol. 28, pp. 1645-1656, 2013.

[56] M. E. Baran and I. M. El-Markabi, "A Multiagent-Based Dispatching Scheme for

Distributed Generators for Voltage Support on Distribution Feeders," Power Systems,


[57] M. E. Elkhatib, R. El-Shatshat, and M. M. A. Salama, "Novel Coordinated Voltage

Control for Smart Distribution Networks With DG," Smart Grid, IEEE Transactions on,

vol. 2, pp. 598-605, 2011.

[58] D. Divan, R. Moghe, and A. Prasai, "Power Electronics at the Grid Edge : The key to

unlocking value from the smart grid," Power Electronics Magazine, IEEE, vol. 1, pp.

16-22, 2014.

[59] Scottish and Southern Energy, "technical Appendix 12 - Making Innovation Happen ",

2013. [Online]. Available:

http://www.energyinnovationcentre.com/wp-content/uploads/SSEPD-Innovation-Strate

gy.pdf. [Accessed: 11-Oct-2015].

[60] C. Goodhand. "Ofgem - LCNF Tier 1 project: 33kV Superconducting Fault Current

Limiter", 2015. [Online]. Available:

https://www.ofgem.gov.uk/ofgem-publications/94151/33kvsfclfinalclosedownreportver

10final-pdf. [Accessed: 11-Oct-2015].

[61] UK Power Networks, " Ofgem - LCNF Tier 2 project: Flexible Urban Networks Low

Voltage", 2014 [Online]. Available:

http://innovation.ukpowernetworks.co.uk/innovation/en/Projects/tier-2-projects/Flexibl

e-Urban-Networks-Low-Voltage/Project-Documents/Overview+Flexible+Urban+Netw

orks+Low+Voltage-+Jan+2014.pdf. [Accessed: 11-Oct-2015].

[62] M. McGranaghan, "Intelligent Universal Transformer Design and Applications",

Jun-2009. [Online]. Available:

http://www.cired.net/publications/cired2009/main_sessions/Session%201/Main%20Ses

sion%201%20pdfs/Block%201/S1%201032.pdf. [Accessed: 11-Oct-2015].

[63] D. J. Rogers and T. C. Green, "An Active-Shunt Diverter for On-load Tap Changers,"

Power Delivery, IEEE Transactions on, vol. 28, pp. 649-657, 2013.

[64] G. L. A. Ghosh, Power Quality Enhancement Using Custom Power Devices: Springer,

2002.

[65] Y. Pal, A. Swarup, and B. Singh, "A Review of Compensating Type Custom Power

Devices for Power Quality Improvement," in Power System Technology and IEEE

http://www.ofgem.gov.uk/ofgem-publications/94151/33kvsfclfinalclosedownreportver10final-pdf

http://www.ofgem.gov.uk/ofgem-publications/94151/33kvsfclfinalclosedownreportver10final-pdf

http://innovation.ukpowernetworks.co.uk/innovation/en/Projects/tier-2-projects/Flexible-Urban-Networks-Low-Voltage/Project-Documents/Overview+Flexible+Urban+Networks+Low+Voltage-+Jan+2014.pdf



http://www.cired.net/publications/cired2009/main_sessions/Session%201/Main%20Session%201%20pdfs/Block%201/S1%201032.pdf

http://www.cired.net/publications/cired2009/main_sessions/Session%201/Main%20Session%201%20pdfs/Block%201/S1%201032.pdf

Reference

145

Power India Conference, 2008, pp. 1-8.

[66] Scottish and Southern Energy, "OFGEM - LCNF Tier 1 Close-Down Report",

28-Feb-2013. [Online]. Available:

https://www.ofgem.gov.uk/ofgem-publications/45826/sset1002-lv-monitoring-lcnf-t1-c

lose-down-report-130228.pdf. [Accessed: 11-Oct-2015].

[67] J. M. Bloemink and T. C. Green, "Increasing distributed generation penetration using

soft normally-open points," in IEEE Power and Energy Society General Meeting, 2010,

pp. 1-8.

[68] J. M. Bloemink and T. C. Green, "Increasing photovoltaic penetration with local energy

storage and soft normally-open points," in IEEE Power and Energy Society General

Meeting, 2011, pp. 1-8.

[69] E. Romero-Ramos, et al., "Assessing the loadability of active distribution networks in

the presence of DC controllable links," Generation, Transmission & Distribution, IET,

vol. 5, pp. 1105-1113, 2011.

[70] J. M. Bloemink and T. C. Green, "Benefits of Distribution-Level Power Electronics for

Supporting Distributed Generation Growth," Power Delivery, IEEE Transactions on,

vol. 28, pp. 911-919, 2013.

[71] N. Okada, M. Takasaki, H. Sakai, and S. Katoh, "Development of a 6.6 kV - 1 MVA

Transformerless Loop Balance Controller," in IEEE Power Electronics Specialists

Conference, 2007, pp. 1087-1091.

[72] M. Noroozian, L. Angquist, M. Ghandhari, and G. Andersson, "Use of UPFC for

optimal power flow control," Power Delivery, IEEE Transactions on, vol. 12, pp.

1629-1634, 1997.

[73] M. Takeshita and H. Sugihara, "Effect of fault current limiting of UPFC for power flow

control in loop transmission," in IEEE Transmission and Distribution Conference and

Exhibition 2002: Asia Pacific., 2002, pp. 2032-2036.

[74] M. Barragan, et al., "Operational benefits of multiterminal DC-links in active

distribution networks," in IEEE Power and Energy Society General Meeting, 2012, pp.

1-6.

[75] G. Fang and M. R. Iravani, "A Control Strategy for a Distributed Generation Unit in

Grid-Connected and Autonomous Modes of Operation," Power Delivery, IEEE


[76] N. Okada, "Verification of Control Method for a Loop Distribution System using Loop

Power Flow Controller," in IEEE Power Systems Conference and Exposition, 2006, pp.

2116-2123.

[77] A. Marano-Marcolini, E. Romero-Ramos, A. Gomez-Exposito, J. M. Maza-Ortega, and

J. L. Martinez-Ramos, "Enhancing the integration of renewable sources in distribution

systems using DC-links," in IEEE Sustainable Alternative Energy (SAE) Conference,

2009, pp. 1-5.

[78] M. A. Sayed and T. Takeshita, "All Nodes Voltage Regulation and Line Loss

Minimization in Loop Distribution Systems Using UPFC," Power Electronics, IEEE


[79] J. M. Maza-Ortega, et al., "Voltage source converter-based topologies to further

http://www.ofgem.gov.uk/ofgem-publications/45826/sset1002-lv-monitoring-lcnf-t1-close-down-report-130228.pdf

http://www.ofgem.gov.uk/ofgem-publications/45826/sset1002-lv-monitoring-lcnf-t1-close-down-report-130228.pdf

Reference

146

integrate renewable energy sources in distribution systems," Renewable Power

Generation, IET, vol. 6, pp. 435-445, 2012.

[80] R. A. A. de Graaff, J. M. Myrzik, W. L. Kling, and J. H. R. Enslin, "Series Controllers

in Distribution Systems - Facilitating Increased Loading and Higher DG Penetration,"

in IEEE Power Systems Conference and Exposition, 2006, pp. 1926-1930.

[81] R. A. A. de Graaff, J. M. A. Myrzik, W. L. Kling, and J. H. R. Enslin, "Intelligent

Nodes in Distribution Systems - Optimizing Steady State Settings," in Power Tech,

2007 IEEE Lausanne, 2007, pp. 391-395.

[82] N. Flourentzou, V. G. Agelidis, and G. D. Demetriades, "VSC-Based HVDC Power

Transmission Systems: An Overview," Power Electronics, IEEE Transactions on, vol.

24, pp. 592-602, 2009.

[83] S. G. Johansson., G. A. Splund., R. Jansson., and E. A. Rudervall., "Power system

stability benefits with VSC dc-transmission systems," in CIGRE Paris, 2004, pp. 1-8.

[84] B. Jacobson, Y. Jiang-Hafner, G. A. P. Rey, M. Jeroense, A. Gustafsson, and M.

Bergkvist, "HVDC with Voltage Source Converters and Extruded cables for up to +/-

300 kV and 1000 MW," in CIGRE Sessions, Paris, 2006, pp. 1-8.

[85] N. R. Chaudhuri, R. Majumder, B. Chaudhuri, and P. Jiuping, "Stability Analysis of

VSC MTDC Grids Connected to Multimachine AC Systems," Power Delivery, IEEE


[86] J. M. Espi Huerta, J. Castello-Moreno, J. R. Fischer, and R. Garcia-Gil, "A

Synchronous Reference Frame Robust Predictive Current Control for Three-Phase

Grid-Connected Inverters," Industrial Electronics, IEEE Transactions on, vol. 57, pp.

954-962, 2010.

[87] I. J. Balaguer, L. Qin, Y. Shuitao, U. Supatti, and P. Fang Zheng, "Control for

Grid-Connected and Intentional Islanding Operations of Distributed Power

Generation," Industrial Electronics, IEEE Transactions on, vol. 58, pp. 147-157, 2011.

[88] L. G. B. Rolim, D. R. da Costa, and M. Aredes, "Analysis and Software

Implementation of a Robust Synchronizing PLL Circuit Based on the pq Theory,"

Industrial Electronics, IEEE Transactions on, vol. 53, pp. 1919-1926, 2006.

[89] S. Cole, J. Beerten, and R. Belmans, "Generalized Dynamic VSC MTDC Model for

Power System Stability Studies," Power Systems, IEEE Transactions on, vol. 25, pp.

1655-1662, 2010.

[90] B. Silva, C. L. Moreira, H. Leite, and J. A. Lopes, "Control Strategies for AC Fault

Ride Through in Multiterminal HVDC Grids," Power Delivery, IEEE Transactions on,

vol. 29, pp. 395-405, 2014.

[91] H. Nikkhajoei and R. H. Lasseter, "Microgrid Protection," in IEEE Power Engineering

Society General Meeting, 2007, pp. 1-6.

[92] N. T. Stringer, "The effect of DC offset on current-operated relays," Industry

Applications, IEEE Transactions on, vol. 34, pp. 30-34, 1998.

[93] M. E. Baran and F. F. Wu, "Optimal sizing of capacitors placed on a radial distribution

system," Power Delivery, IEEE Transactions on, vol. 4, pp. 735-743, 1989.

[94] G. Daelemans, K. Srivastava, M. Reza, S. Cole, and R. Belmans, "Minimization of

steady-state losses in meshed networks using VSC HVDC," in IEEE Power & Energy

Reference

147

Society General Meeting, 2009, pp. 1-5.

[95] M. J. D. Powell, "An efficient method for finding the minimum of a function of several

variables without calculating derivatives," The Computer Journal, vol. 7, pp. 155-162,

1964.

[96] L. Cheng-Ling, C. Tai-Ning, L. Yuan-Yao, H. Pin, and R.-K. Lee, "Powell's method for

designing optical multilayer thin-film filters," in 15th OptoeElectronics and

Communications Conference, 2010, pp. 388-389.

[97] S. Lazarou, V. Vita, and L. Ekonomou, "Application of powell's optimisation method

for the optimal number of wind turbines in a wind farm," Science, Measurement &

Technology, IET, vol. 5, pp. 77-80, 2011.

[98] H. Cho, A. D. Smith, and P. Mago, "Combined cooling, heating and power: A review of

performance improvement and optimization," Applied Energy, vol. 136, pp. 168-185,

2014.

[99] M. E. Baran and F. F. Wu, "Network reconfiguration in distribution systems for loss

reduction and load balancing," Power Delivery, IEEE Transactions on, vol. 4, pp.

1401-1407, 1989.

[100] UK Statutory Instruments, No. 2665, “The Electricity Safety, Quality and Continuity

Regulations,” 2002.

[101] M. A. Azzouz, M. F. Shaaban, and E. F. El-Saadany, "Real-Time Optimal Voltage

Regulation for Distribution Networks Incorporating High Penetration of PEVs," Power

Systems, IEEE Transactions on, vol. PP, pp. 1-12, 2015.

[102] A. E. Kiprakis and A. R. Wallace, "Maximising energy capture from distributed

generators in weak networks," Generation, Transmission and Distribution, IEE

Proceedings-, vol. 151, pp. 611-618, 2004.

[103] "IEEE Application Guide for IEEE Std 1547(TM), IEEE Standard for Interconnecting

Distributed Resources with Electric Power Systems," IEEE Std 1547.2-2008, pp.

1-217,

Publications

148

Publications

Journal Paper:

[1] W. Cao, J. Wu, C. Wang, T. Green and N. Jenkins, “Operating Principle of Soft

Open Points for Electrical Distribution Network Operation” Applied Energy, Vol.

164, pp 245-247, Feb. 2016. 10.1016/j.apenergy.2015.12.005

[2] W. Cao, J. Wu, C. Wang, T. Green and N. Jenkins, “Benefits Analysis of Soft Open

Points for Electrical Distribution Network Operation” Applied Energy, Vol. 165, pp

36-47, Mar. 2016. 10.1016/j.apenergy.2015.12.022

Conference Paper:

[1] W. Cao, J. Wu, and N. Jenkins, “Feeder Load Balancing in MV Distribution

Networks Using Soft Normally-Open Points," in IEEE PES Innovative Smart Grid

Technologies (ISGT), Europe, 2014, PP 1-6.

[2] W. Cao, J. Wu, and Z. Huang, “Soft Open Points for Supply Restoration in

Medium Voltage Distribution Networks” accepted for publication in IEEE PES

Asia-Pacific Power and Energy Engineering Conference (APPEEC), Australia,

2015.

[3] A. Aithal, C. Long, W. Cao, J. Wu, “Impact of Soft Open Point on Feeder

Automation” accepted for publication in IEEE ENERGYCON, Europe, 2016.

http://dx.doi.org/10.1016/j.apenergy.2015.12.005

http://dx.doi.org/10.1016/j.apenergy.2015.12.005

Appendix A

149

Appendix A: Data for the Two-Feeder Test

Network

Table A.2 Network Parameters for the SOP study in Chapter 5

Feeder

No.

From

bus

To

bus

Line Impedance parameters Length

(km)

load parameters at

receiving buses

r (ohm/km) x (ohm/km) P (kW) Q (kVar)

2 0 1 0.203 0.1034 1 60 25

2 1 2 0.2842 0.1447 1 60 25

2 2 3 1.059 0.9337 1 60 20

2 3 4 0.8042 0.7006 1 120 70

2 4 5 0.5075 0.2585 1 200 600

2 5 6 0.9744 0.963 1 150 70

2 6 7 0.3105 0.3619 1 210 100

2 7 8 0.341 0.5302 1 60 40

1 0 1 0.0922 0.047 1 100 60

1 1 2 0.493 0.2511 1 90 40

1 2 3 0.366 0.1864 1 120 80

1 3 4 0.3811 0.1941 1 60 30

1 4 5 0.819 0.707 1 60 20

1 5 6 0.1872 0.6188 1 200 100

1 6 7 0.7114 0.2351 1 200 100

1 7 8 1.03 0.74 1 60 20

1 8 9 1.044 0.74 1 60 20

1 9 10 0.1966 0.065 1 45 30

1 10 11 0.3744 0.1238 1 60 35

1 11 12 1.468 1.155 1 60 35

1 12 13 0.5416 0.7129 1 120 80

Appendix A

150

1 13 14 0.591 0.526 1 60 10

1 14 15 0.7463 0.545 1 60 20

1 15 16 1.289 1.721 1 60 20

1 16 17 0.732 0.574 1 90 40

Appendix B

151

Appendix B: Data for the Four-Feeder Test

Network

Table B.3 Network Parameters for the Case study in Chapter 5

Feeder

No.

From

bus

To

bus

Line Impedance parameters Length

(km)

Rated loading at

receiving buses

r (ohm/km) x (ohm/km) P (kw) Q (kvar)

1 0 1 0.55 0.4 1 228 171

1 1 2 0.55 0.4 1 273.6 79.8

1 2 3 0.55 0.4 1 112.5 124.26

1 3 4 0.55 0.4 1 171 228

1 4 5 0.55 0.4 1 79.8 273.6

1 5 6 0.55 0.4 1 256.5 124.26

1 6 7 0.55 0.4 1 228 171

1 7 8 0.55 0.4 1 163.2 68.6

2 0 1 0.55 0.4 1 196 147

2 1 2 0.55 0.4 1 235.2 68.6

2 2 3 0.55 0.4 1 220.5 106.82

2 3 4 0.55 0.4 1 147 196

2 4 5 0.55 0.4 1 68.6 235.2

2 5 6 0.55 0.4 1 235.2 68.6

2 6 7 0.55 0.4 1 196 147

2 7 8 0.55 0.4 1 235.2 68.6

2 8 9 0.55 0.4 1 196 147

2 9 10 0.55 0.4 1 343.2 100.1

3 0 1 0.55 0.4 1 286 214.5

3 1 2 0.55 0.4 1 321.75 155.87

3 2 3 0.55 0.4 1 307.2 100.1

Appendix B

152

3 3 4 0.55 0.4 1 100.1 343.2

3 4 5 0.55 0.4 1 214.5 286

3 5 6 0.55 0.4 1 55.2 68.6

4 0 1 0.55 0.4 1 196 147

4 1 2 0.55 0.4 1 235.2 68.6

4 2 3 0.55 0.4 1 220.5 106.82

4 3 4 0.55 0.4 1 147 196

4 4 5 0.55 0.4 1 235.2 68.6

4 5 6 0.55 0.4 1 220.5 106.82

4 6 7 0.55 0.4 1 235.2 68.6

4 7 8 0.55 0.4 1 111 196

4 8 9 0.55 0.4 1 106.82 220.5

Soft Open Points for the Operation of Medium Voltage ...s PhD thesis_02.02.2016 (clean... · Soft...

Documents

Transcript of Soft Open Points for the Operation of Medium Voltage ...s PhD thesis_02.02.2016 (clean... · Soft...